The source code repo is still on Epic's Github, not sure if everything else is that easy to get currently. The problem about picking it up by a 3rd party is the fun of getting approval on basically anything.
I compiled the project on a whim during COVID. It played okay. There was even a very small pool of players who had presumably just done exactly the same thing.
Nope, piece selection in the unmodified NEStris uses only the player input to generate pieces and it also starts from the same seed, so frame-perfect inputs after loading the game will result in the exact same pieces. The first perfect level 18 run (only tetrises to score) was dismissed by community because it was made by replicating a TAS run with a suitable piece sequence.
Perfect fourth could be achieved via natural harmonics at 5th and 7th frets, double octave is just the 5th fret and voila you can get the standard tuning. Inapplicable for necks without a precise position of frets, good luck compensating height differences from nut and gauges with intonation tuners.
Oh yeah, I should've mentioned fretting the lower strings for a unison (1:1) to the higher adjacent open string. No beating in this unison yields equal temperament, due to frets scaled accordingly, so it avoids the problem I mentioned... but introduces likelihood of intonation issues (or simply bending the pitch) as you say.
I had been thinking about listening for beating in the 5:4 and 4:3 intervals using open strings, which does present the just temperament problem, and harmonics on open strings to make a unison unfortunately shares this problem 100%.
Last time I had a galactic blow-up of apt solver (the final part of 64-bit time transition in Debian Testing) it was mere 2 GiB of memory per minute.