I can't comment specifically on the Anovos TK, since I haven't studied the differences at all, but as a more general explanation I would assume it's down to how the scan was handled afterward. I've worked with 3D scans of props and you can't really use them 'out of the box' so to speak. Scanning can introduce errors and glitches that need to be removed. Scans also usually produce "bad geometry" which is to say that the polygons that make up the surface of the model are arranged in a way that makes them look 'messy' and hard to work with, so you need to clean that up, without destroying the captured details. In this stage it's critical to have a 'good eye' so that you preserve all the details exactly. Sometimes you see something, for example a little 'bump' or whatever, that you can't determine whether it's part of the actual prop or a scan anomaly- that's when you start to pore over photos like a madman so that you don't remove an authentic detail. If someone is less experienced, less meticulous (or just in a hurry) there's a high likelihood they will smooth over certain details or otherwise alter the object, without realizing it. Think of scanning an object as casting it in clay- you can still go in and change things. Again, I don't know much about the Anovos TK, but if they made metal vac bucks, perhaps they had to do lots of alterations to make creating them easier?
No disrespect intended towards whomever worked with the Anovos files, but it's simply possible that you, GF, are more adept at spotting minute differences after studying the TKs for so long. (I am kind of the same way with some props from Guardians of the Galaxy.) To most, the Anovos TK looks like any TK, but to you, every little error is a little pin jabbing your retina. I mean, there are folks out there that can't see much difference between an FX and a direct-lineage pull!
I know (and agree with) the idea that holding a solid object in your hands is different than working from a screen, but I still consider a 3d scan a direct lineage, because instead of using silicone to 'capture' the location data of molecules in the prop, you're using light and a memory card. It's just capturing the information in a different way. In fact, if you're really good at it, scanning an object can create a more exact copy because you don't have to deal with thinks like shrinkage.