Oh, plenty of things. Full disclosure: I did not purchase OOTP 20 so some of this may have been addressed, but based on other posts I've read it doesn't seem so, at least to the point where it would be a worthwhile purchase for me.
I'm a fictional league player but it's basically an MLB environment, so that's where I'm coming from but much of this applies to other game modes to.
1. Statcast-type stats, as somebody posted before. Batted-ball stuff, spin rate, route efficiency, etc. Tie it into scouting and player development. Which brings me to my biggest one...
2. Overhaul the entire player development engine. Make aging curves more realistic... for example, Juan Soto was a star last year at age 19. Even with adjusting player development speed a good deal, I couldn't really replicate things and get some 19, 20, or 21 year old stars.
The way prospects develop and the sort of stats they produce don't seem to correspond to how things are in real life. Look at the top 10 prospects in baseball... guys like Vlad Jr, Tatis Jr, Wander Franco, Forrest Whitley, etc... The guys who would theoretically be 75 or 80 potential players in OOTP. For the most part, these guys dominate the minor leagues. If they're not putting up dominant numbers, then they're putting up respectable numbers at a level for which they're quite young.
I'm not saying it should be the guys with the best potential have the best numbers and vice versa, but if you look at an OOTP top prospects lists the top end guys don't really produce the numbers. When I follow a top pitching prospect in real life, each start they make is exciting to see how they do. In OOTP, it doesn't matter because the stats and the potential don't really relate at all. This is also why I can't do stats-only. I think I'm doing a poor job of explaining it but there's just something "missing" in the whole scouting and player development part of the game. But playing with new scouting reports once a year, I would regularly have young prospects put up good numbers at a challenging level and see their potential rating plummet. Or a guy would have a terrible year but his potential stayed the same.
Maybe instead of just a scouting director we could have a team of three or so scouts? Then instead of just OSA or SD we have some more options? By the way, OSA should be a little bit better than it is. Across all the places that put out Top Prospect lists, there's relative consensus at the top.
3. Add options for alternate uniforms, especially as the 3D becomes more and more a part of the game.
4. This is a tiny one... When you do "Make this work now" on a trade, it produces the list of players. If I open up a player's profile to see if I want to include him in the trade, then go back, I have to click "Make this work now" again to get the list back up.
5. Much more realistic contract demands. I believe this is something that was addressed in OOTP 20 but the way veterans asked for huge deals (and found teams to oblige) was crazy in OOTP 19. A team would shell out a $150 million contract and then by midseason that player would be on the trading block because the team was trying to dump him as an overpaid veteran.
6. Improvements to league history and almanac. Let me get really, REALLY immersed. Easy-to-see playoff brackets from prior years. Sortable draft history for the league and for teams similar to baseball-reference. Which first-round picks in league history have produced the most WAR? Stuff like that.
7. Save lineups for my main league team when players "leave" the league to play in a tournament like the World Baseball Classic.
That's what I can think of for now. I'll post more I'm sure... I should note that I'm still playing OOTP 19 more than a year later because it's a very, very good and fun game. Don't want that to get lost when I'm basically posting a list of complaints
