Perhaps I'm overreacting, but I can't see being able to 100% positively ID any critter at 500 yards with any 35mm thermal smaller than an adult cow.
Playing devil's advocate here, if you are having that much trouble, then I am not sure how you will know it is an adult cow. The identification process varies from person to person. I make a lot of very long distance "identifications" based on shape and movement. These are 2D shapes, right, which is the nature of thermal. These are not quick identifications and sometimes may require minutes to do. I suspect that other folks do the same thing. One could argue that I am not making proper identifications and I would agree as a matter of fact, but not as a matter of practicality. Moreover, I would argue that everybody using thermal except at close range is doing what I do or a variation of it in order to make their PIDs and feel completely justified in their minds when using thermal. Calling them "identifications" or "PIDs" may be a bit of a misnomer.
Except as close ranges, people are going through a decision tree process of identification to rule in and rule out targets. Case and point, most folks would not be able to rule between a blackbacked jackal and a coyote at 200 yards through thermal. The counter argument to that is, "But there are no blackbacked jackals running free in the United States. Okay. That may be true, but given that it is true, can you truly identify the coyote at 200 yards or are you making a series of assumptions such as that nobody has released some blackbacked jackals or similar animals? Could you tell them apart at 100 yards? Does the American hunter even know the profiles of all of the potential canids well enough to make such distinctions. If not, then is an "identification" truly being made or is it just a recognition and best guess given likelihoods? It is the latter.
My point is that we are all makes a series of assumptions when we make identifications at range with thermal, even with a lot of pixels.
How good one becomes at interpreting their pixelated images will go a long way toward their success and failure. I find that I can "identify" targets much better than folks with less experience when using the same gear. Things I often see being confused are hogs and deer (with heads down in tall grass), coyotes and deer, jackrabbits and coyotes, hogs and bedded calves, feral hogs and javelina, calves and deer, raccoons and opossums, raccoons and armadillos, opossums and armadillos, PSRs (pig-shaped rocks) for hogs.
So I am not knocking what you are saying, just including the aspect that a lot of PIDs are anything but truly PIDs. They are assumptive, based on interpretation and experience which will play into how accurate they turn out to be.
