Brainstorming is a creative process that we encourage as jams are a good time to try something new or test your skills, but trying to decide what game you'll make is going to have deminishing returns due to the theme being an important part of this jam.
Games should be made during the jam. But logistics is pretty fair game: for example, personally I like to figure out if I'm going to be working in 2D or 3D, the engine I'll use, team assembly, the version of the engine the game will be made on, what source control is needed for the team (github), any asset packs that might be helpful in a pinch, how the team is communicating, time zones, discord servers, trello boards, file sharing, --- and.. in the past I've also decided certain genre/mechanics with teams we'd find interesting to make, it's just fun to say "hey no matter what let's make it multiplayer" but we don't hook anything up until the jam starts. Teammates should be assembled ahead of time and you should all be ready to start when the clock starts, but nothing about the game itself should be started.
Brainstorming, imo is going to yield the best results when the theme is revealed as you'll build a game around the theme mechanically.
Hope that helps!