Thank you very much!
My goal for the piece selection was to find a set with similar character or function to our familiar seven tetromino friends. (In playtesting, I even experimented with getting the piece count up to six with strictly distinct mini-L and mini-J pieces with two rotation states each, but it ended up being important to unify them into a singular four-state piece to deal with how much of a villain the "T" can be.)
Thank you for the feedback on the speedcurve. I'm going to take another look at the gravity table to address the pace of the early game. I think in trying to focus on adjusting the mid/late game speed within the "alternate history design and programming practice" framing I had for the game, I had overlooked the slow speed of the early game. I can tweak things to get a better fit to what I was aiming for.
The playfield is juuust a bit "smaller than regulation." It's 9 wide to fit three 3-wide pieces cleanly. Keeping the conventional 2:1 height/width ratio, pieces now spawn at a height of 18; this does result in pieces landing at a slightly brisker pace at the same level of gravity, similar to the gravity disparity between GB Tetris and NES Tetris (row 17 spawn vs. row 20 spawn, owing to the former's reduced vertical screen resolution). I did consider experimenting with even shorter or narrower playfields, but ultimately decided it felt about right as-is.