Well you got what you wanted John, this is certainly an interesting discussion.

Obviously without any documentation or markings we will never know, but personally I think that you have given more reasons for it to be a possible Harrington Gat prototype than those simply dismissing the idea.

“The absence of evidence is not the evidence of absence.”

Just because there are no markings or documentation doesn't mean it wasn't made by T.J.Harrington

Along with the fine level of trigger adjustment I think the spring being in three well finished sections is a strong sign of an intention to experiment with different setups, and in fact if the power was to be changed with different spring section combinations then the trigger weight would have to be changed accordingly to give a safe amount of sear engagement in relation to the strength of the spring. Not something that someone just making a project gun would probably consider...
Maybe having to remove the sideplates to adjust the trigger meant that the user could get a visual impression of sear engagement etc. while adjusting the trigger, which may have been more useful for experimentation than simply adjusting the trigger blindly from the outside? Just an idea.

It's difficult when we see what we want to see, and it can be a bit delusional to jump to convenient conclusions about an airguns unknown history or connections to well known later models....I don't think that is the case here though.
You've earned the right to make a judgement on the rough age of the patina, and if it does predate your pre war Gat then the similarities are too much of a coincidence, and if it's post war and after the Gat was introduced then why would anyone bother painstakingly copying a cheap existing pistol? (Aero roofracks aside)

All we can do when investigating unknown origins of these old airguns with no provenance is build a case for a probable explanation, with evidence where possible, and then conclude a percentage of probability that we wish to believe.
Personally given everything you've put forward I think your 60% probability of it being a prototype is perfectly reasonable, it's not as if you're saying that it definitely is the case. That would be as unreasonable as someone saying that it definitely isn't.


Cheers,
Matt