Well, I have to admit it.
It was tempting to draw large tyrannosaurs with a heavy coat of feathers. It proved VERY tempting. But I stuck to my guns and held out on it. Why? I wasn't convinced the evidence would favor it.

I had already seen pictures of skin impressions attributed to the neck and chest of T. rex, and they were scaly. But this wasn't published and therefore people were casting doubt on it (I wonder if they did the same when Yutyrannus photos showed feathers, before it was formally published and described?). Now some may ask what I have against feathered tyrannosaurids, since that is the new orthodoxy in much of paleo-art (just as lizard-like restorations were orthodoxy in the time of Knight and Burian). The answer is actually: nothing. But when talking about giant tyrannosaurids there was no actual evidence of feathered skin. Not only that, but being roughly 1.5 times the length of the still taxonomically controversial Yutyrannus, the largest T. rexes at 12 meters were over twice the mass, and in addition they were living in a time with warmer global temperatures and increased SO2 and volcanic activity.

I maintained for years that if T. rex had any feathers as an adult, they would probably be in very limited areas, for display. Something that big would not need them for insulation, and as a warm-blooded predator in the Maastrichtian epoch with rising global temperatures, it would likely overheat with a heavy coat of downy insulation feather, and obviously had no use for long-quilled flight feathers.

Now we have a paper just out confirming that big tyrannosaurids were probably scaly over the majority if not all of their body surface - and that feathers were lost relatively early in the evolution of true Tyrannosauridae.
And this isn't some fringe paper in a fake journal by creationists or BANDits. It's Royal Society, and Phil Currie, Bob Bakker, Darren Tanke and Pete Larson are all among the co-authors. You don't get more solid in credentials than that. I don't even think Jack Horner would take issue with their scaly conclusion, given his historically conservative approach to dinosaur biomechanics and metabolism.

Interestingly, not only are the skin impressions scaly, but the scales are tiny. Bead-like, almost. So most likely these are from a softer part of the body, possibly the undersides, but due to decomposition the skin may have become desiccated or detached, and rolled around to the dorsal portions of the body, which are apparently the positions in which the skin impressions were found.

Now I can imagine some people having nightmares over their entire post-dotcom idea of a poofy T. rex being overturned. So what does this really mean for the big picture of tyrannosaurs?

Does it suddenly mean they were cold-blooded overgrown alligators? NO. Warm-bloodedness is not dependent on feathers or fur, especially not at those sizes.
Does it mean that they were unrelated to birds and we need to rip up the theropod family tree? NO. The paper states that the tyrannosaur lineage lost their feathers, not that they never had them.
Does it mean that all the old pre-Bakker images of sluggish tail-dragging T. rexes were correct? NO and no. These certainly weren't the only unfeathered tyrannosaurs ever drawn or painted, and definitely not the most anatomically informed (Burian was even famous for not looking at fossils or estimating proportions). There's no reason to throw out 40 years of gathered evidence of warm-blooded, fast-growing bone texture in tyrannosaurs. Tyrannosaurs were still active, fast, warm-blooded hunters. They just weren't drowning in feathers at 40 feet long.
But this doesn't mean that I favor Jurassic Park's ugly monsters either. And simply favoring scaly T. rexes doesn't make you some obsessive JP fanboy. It simply means you listen to the evidence - several scaly impressions in 5 different tyrannosaur genera, and not one feathery impression from any of them. Interestingly, none of these impressions match the big, bulky crocodile-like scales and heavy wrinkles in the Jurassic Park T. rex. Yet I already see and hear some hurt naysayers claiming that everyone who questions the unproven notion of T. rex feathers is somehow a Jurassic Park fan with an agenda who hates feathers. It couldn't be further from the truth.

I was never a big fan of the Jurassic Park franchise, and even now the scale evidence doesn't support their big alligator-like scales or heavy baggy wrinkles on their T. rexes. What's more, the JP 'rexes had obvious errors in their facial appearance, including the lumpy 'gator-like orbital horns and huge triangular jugals, and the "cutout" sections of the lower jaw that they fit into... As if somehow having any overlap of the upper jaw around the lower was a sin to Spielberg, as if they had to close perfectly interlocked like a crocodile. We know that the scale impressions show much smaller and finer scales than anything in the JP franchise. My reasons for questioning the idea of giant 8-ton tyrannosaurs being feathered have nothing to do with a hollywood movie. And everything to do with (a) the evidence, and (b) the simple fact that they were giant 8-ton tyrannosaurs in a rapidly warming Hell Creek climate.

So they did have scales over several parts of the body, and the scales in the hip area indicate it's highly likely the torso was scaly as well. You just don't see scaly hips in any feathered creature, even flightless birds have feathers on the hips and tail (or what's left of it). Of course this doesn't change anything about dinosaurs that we know either had feathers or were likely to have them - like dromaeosaurids ("raptors"), ornithomimids, oviraptorids, caegnathids, etc. A scaly tyrannosaur doesn't negate these animals being feathered, nor does it have any effect on feathered specimens of controversial species sometimes labeled as basal tyrannosauroids, but which may not be (i.e. Yutyrannus).

In fact, the paper's argument is that tyrannosaurids evolved from small feathered ancestors, but likely later lost the feathers as they evolved into giant top-tier predators.

In that case, knowing that the scales were very small, and nothing like the baggly wrinkles and 'gator scutes in Jurassic Park, we can now ask: what is the most accurate T. rex image, based on these scale impressions? It would obviously have to be something post-70s, when Bakker and the Dinosaur Renaissance were gaining steam, and anatomical accuracy really began to matter.

You might say Greg Paul's version. And you'd have a point. Though many of the visible scales here look a good deal larger than the ones in the impressions, not all the scales are visible, and some patches on the underside look like a good place for much finer scales. If the rest of the body was covered in similar, if somewhat larger scales, you would likely have the smooth, tight, and only lightly wrinkled skin you get on Greg Paul dinosaurs.

T. rex pair by Greg Paul. 1988. Based on the AMNH specimen. Used for educational purposes only.

However, Brian Franczak's slightly less "vacsuit" version is another contender:

Brian Franczak's T. rex and Edmontosaurus. 1991. Used for educational purposes only.

There's not much surface detail in Franczak, aside from a few veins and light wrinkles. That's just the style. But overall the anatomy looks solid here and the skin smooth enough that those tiny scales in the impressions would be right at home. The legs are a bit more bulky here, but still not like in JP. The head looks oversized in the far left animal, but that's just a perspective trick for dramatic effect.

Another great rendition of accurate, fine-grained T. rex skin is in Larry Felder's painting of a juvenile 'rex, which was the cover image for the book In the Presence of Dinosaurs. You can almost feel this is a real live animal breathing next to you. The scales struck most people as too tiny, the skin too smooth and "cute" - but after seeing these skin impressions with some scales only a couple of millimeters wide, it's hard not to wonder if Felder had some kind of sixth sense for these things.

Larry Felder's young T. rex from the book cover. Not based on a particular specimen. Educational purposes only.

But if you were to ask me which painting of T. rex looks the most like the actual textures and scale shapes in the skin impressions, I'd have to give it to this long-forgotten gem:

T. rex by Dave Marrs. Apparently based on the Wankel/MOR specimen. Used for educational purposes only.

Now I don't know how many of you are familiar with Dave Marrs' work. He did a large number of paintings for the TV series PaleoWorld in the 90s, which is where I was first introduced to the art of Greg Paul, Brian Franczak, Mark Hallett, David Peters and many others (back then Peters was solidly popular as an artist, and followed the best cutting-edge science of the day, almost like a second Greg Paul - and there wasn't anything too unusual about his paleo-images at that time, even the pterosaurs). Marrs also did some unrelated images for a series of Jurassic Park trading cards, which were largely concept art for the movies.

Apparently Marrs got into legal hot water when he closely copied a number of images from Paul and Peters, robbing both to pay himself and not getting permission to copy copyrighted images. Some of these "re-skinned" copies were also featured in PaleoWorld, most notoriously his Estemmenosuchus, which was a ripoff of Peters' original. But, it must be emphasized, not everything Marrs did was a fraud. He also had many original paintings and no doubt he proved his skill as an artist.

Probably his best painting was the one above of a blue T. rex with relatively thin, basal lips and some VERY detailed scalework. The facial scales are a bit on the large side, but those on the underside of the neck are downright tiny. And this matches the look of the skin impressions almost perfectly. Now given, there are some flaws with the oversized teeth and the jaw proportions, but some of this is due to perspective distortion and stylistically it's a very good painting. And if the skin impressions are any clue, this is easily the most accurate skin in a T. rex painting. Too bad more isn't known about this image, it really does stick with you.

From a distance, some vintage "Dinosaur Renaissance" paintings, like Greg Paul's Daspletosaurus, even did away with extensive scale texturing on the majority of the skin altogether, leaving it to the viewer to imagine what those tiniest of scales looked like up close.

Daspletosaurus by Greg Paul. Based on holotype CMN 8506. Updated version with revised supinated hands. Used for educational purposes only.

Now in the future someone might find skin impressions in other parts of the body of a tyrannosaurid, that indicate the presence of feathers. But so far there are none, and 100% of the few rare skin imprints found of tyrannosaurs thus far, are scaly. The trend to put feathers on tyrannosaurs was always a speculative one, with no actual tyrannosaur evidence to back it up, the closest analogues used being only distantly related and much smaller theropods like Guanlong, Dilong, the maniraptora, and the possibly non-tyrannosauroid Yutyrannus, whose size has often been overestimated. The new paper shows that the five best-known tyrannosaurid genera were scaly over at least several parts of the body, and it's unusual to have feathers limited only to the torso or other regions lacking skin impressions. Even flightless birds today don't have naked hips or tails.

Which brings up an important thought - perhaps the scaly but nevertheless sleek and hot-blooded "Dinosaur Renaissance" T. rexes of the '80s and '90s are not so far off the mark after all. Aside from the pronated hand postures common to that era, they pretty much got it right.