Most of this list can easily be explained by lumping them into 2 categories.
1) Most of them fit here. Anything that isn't represented in a basic newspaper box score significantly limits the player pool. Part of the beauty of the way this game is designed, as opposed to most other historical replay/mix-and-match games on the market, is that it works purely off of the real historical statistical record of Major League Baseball. This is the same reason why realistically negro league players can never be added to the SLB database. Many other games convert stats to ratings that, if the player plays in a league of similar talent and statistical makeup to his historical league, should reasonably reproduce that player's stats over time. SLB doesn't do that. It uses exclusively statistical inputs. That means if we want to use players from the early eras of baseball, we have to limit what skills can be used in the game. Successful bunts, for example, generally show up in box scores. Failed bunt attempts do not. This is related to the biggest problem with early-era players in SLB - caught stealing were not always tracked faithfully, so stolen base rates and catcher CS%s for deadball players are iffy at best. And periodically SLB players do complain about this. I don't think we want to exacerbate the problem by trying to extrapolate/invent more numbers where they don't exist. To come back to the example of bunt success rates - any meaningful estimate would require you to know how many foul bunts a player laid down in a season. To know that, you need not just play-by-play accounting, but pitch-by-pitch. That doesn't exist for every game until well into the 21st century.
2) Sample sizes. A lot of things just don't come anywhere close to statistical significance within the statistical framework of a single season, which is the basis for the 99%+ of SLB leagues that aren't DL career leagues. Incorporating additional statistically insignificant numbers and trying to pretend they're meaningful is going to break the game. Consider the case of "clutchness." Without going down the rabbit hole of the ongoing argument about whether baseball players can really be "clutch," it's clear that even those who do believe in clutchness as a skill probably shouldn't want to see it incorporated into this game. Last year Aaron Hicks went 11 for 28 in late and close situations. Which do you think is realistically a better representation of his ability to hit - the .398 he hit in 28 at-bats, or the .235 he hit for the season? Consider that in 2018 he went 12 for 71 in late and close situations, for a .169 average. And he hit .310 late and close in 2017, and .208 in 2016. So did he really go from being poor, to good, to really bad, to really good as a clutch hitter over the past 4 seasons? Or does it have more to do with the fact that, according to most estimates, batting average doesn't really stabilize until you have about a season and a half of full time play's worth of at-bats? Given that, is there really any way in which adding any small sample size tendencies could possibly make the game better, rather than just more random/extreme?
In addition to all of the above, my personal opinion is that the current complexity of the game "feels" about right to me. Adding a bunch of additional statistical depth would probably just make the game less approachable. Most players already struggle with their first few teams. I don't think the way to grow the game is to make the barrier to success taller. Even for me, as a vet, I don't want to deal with a great deal more to think about during drafting. Sure, we could use the same questionable level of accuracy we have for stolen bases to include pitchers in the equation. While this is more realistic, it's also more mathematically complicated, because you're trying to disentangle the pitchers' and catchers' contributions to stolen base rates and success. But you could come up with an equation to do this. You also run into sample size issues with relief pitchers in particular. Even if we wanted to ignore all that, I just personally don't want to have to think about this 3-way interaction during drafting. I'm sure everyone's personal tastes differ a little bit on this question, but while I recognize how frustrating it is to watch your Andy Pettitte get lit up on the basepaths, I still find it preferable to having another thing to think about when drafting pitchers and catchers.