For our team, it was the fact that we didn't have a minimum viable product ready sooner for others to help us play-test. This meant that we found a number of unprecedented balancing & quality-of-life issues rather late in the jam, and some of them couldn't be resolved before the submission deadline.
As an example, some of the minigames in It's Rude To Point didn't explain the objective clearly enough, which was problematic considering how little time we gave players to react. One way we tried to circumvent this was to give objects differently-colored outlines, which would indicate whether an object needed to be grabbed or flicked. Though it was implemented without any issues, we never really explained it in-game, instead bargaining on the notion that perhaps the player could figure it out themselves. Since there was only one minigame involving grabbing, however, I can definitely understand how some wouldn't have quite caught on.
I guess one could argue that participating in a jam is about forming an MVP in as little time as possible, but I think our game could've really benefited from some more polishing and tweaking. That being said, I'm still very proud of what my friends and I managed to accomplish in just 48 hours!
If you'd like, you can check out our entry here:
https://itch.io/jam/gmtk-2019/rate/461113