1) Yes, "down load" is the term of art for loading fewer in a magazine than the rated capacity. "Down load by 2" is common for M16/M4, so the crosscheck is the same as for fully loaded then you can do the same thing no matter who loaded your mags.
2) Yes to much of the Thompson Blish lock: it was before really good analytical or observational methods. The lock absolutely works in principle and at higher pressures, etc. He just didn't do it right on the .45. I always wanted more to the story so we could find why he (apparently) didn't at least pull an Edison (or Dyson) and iterate a few dozen or hundred times. Say, drill the bolt out to lighten it and fire in a test fixture progressively, see when velocity gets too high. He'd have immediately noticed the lock isn't working, and adjusted it and regardless have better optimized the mass / spring for the bolt. Luck was in the end on his side.
3) Have been using a CMMG radial upper in IDPA and generally messing about for a while. I don't track except on my precision rifle, but got to be 1,000 rounds through it. Dead reliable (like an AR 15, and very much unlike the straight blowback Colt SMG, never need to clean it), feels to the shoulder the same as any pistol carbine, but splits are MUCH faster, so it works. A lot. Totally suggested.