"Source" is basically the project's folder you created and used to create the game client's build. For example, with Unity, it's the folder that contains the .vs (hidden), Asset, Library, Logs, obj, Packages, ProjectSettings, UserSettings, .vsconfig and a bunch of .csproj files (and a .sln files with the project's name). To put it in simple terms, it's what's required to build a full working version of your project directly from the engine client itself.
It doesn't have to includes original sources files in foreign formats if those aren't used directly in the engine. For example, files like a .ai (Adobe Illustrator) to makes .svg files or .psd(Photoshop) to make the .png or .jpeg or .tiff files are not necessary. If you use .psd files as-is in the engine (like Unity that allows it), you HAVE to includes those, but if you exported the files as another format to be used, you just have to includes those exported files. It's the same with audio files such as original recordings and files made by particular audio editing software. If you recorded yourself and edited the sounds, you don't have to include the original recording nor the specific files generated by the editing software that contains the editing (and may requires a paid program), but only the result files (.wav, .mp3, .midi).
There's a kind-of-gray area regarding paid tools allowed which allows you to use paid tools as long as you don't have to include said tool in your project. For example, if you can make your soundtrack in Music Maker with paid sounds effects as you don't have to include the Music Maker's file, but only the MP3 (.mp3) or Wave (.wav) files resulted from exporting the soundtrack. This is why they give the example of Gaia which is a paid asset that allows you to generate Terrain Mesh data (doesn't require Gaia to use anymore) as long as you remove Gaia from the project's assets when submitting the sources files while keeping the resulted Terrain Mesh data files.
From what there has been mentioned, all members in a group gets a free course as long as it's within reasonable. Team with 10+ members might be a bit overkilling it unless in a specific situation and you should check the other posts about it. I have always worked alone so I can't tell how to "make" a group for submitting a project on itch.io, but a safe bet would be to include, in your project, a credits list including their itch.io usernames. How the free course will be distributed is yet to be known, except that people will get it after submitting. Maybe they will send a message to the group leader (the one who submit the project) which will allow you to reply a list of GameDev.tv account emails and which course each of those emails can get unlocked.
Just to make it clear, I'm not part of the group managing this Dev Jam. It's my 3rd one. I did one like this one (but a 1-year dev jam which I did within its last 2 months for fun), 1 closed 24-hours Dev Jam (the kind which you CAN'T prepare ahead since you work on someone else PC) and this is my 3rd one. (The main reason why I don't participate in many Dev Jam is because I live in Quebec in Canada and the laws here for contests are managed by a government's group called Loto Quebec which make it relatively impossible for any international contests to allow the it to be run in that province, unless the prize is non-monetary or educational. To put it simple, if the prize is either monetary or of a greater value than 500$ CAD and is not educational and the contest is not to fund a non-lucrative registered event, the contest management has to pay 50% of the 3 main prize values in a fee. If you look at any international contest with such value in prize, you'll often see "Excluding Quebec in Canada" in the condition. As this contest has an educational-based prize, I can participate in it.)
To give an idea of how you can prepare for this kind of open Dev Jam, I'm already done with the general concept of my project and timeline for those 10 days. Considering how long it took me to build up the idea, I saved around 3-4 days of none-productive brainstorming and, at this moment, I got a really good idea of what it will be like. Saving 3-4 days out of 10 days is insanely big and can make an huge difference. I took notes of what kind of menu, inputs, gameplay element the game will have. I also got a few potential ideas depending on what kind of theme will be revealed as my idea can be converted in almost any kind of themes relatively easily. (I do have one favorite kind of setup if the theme allows it, but that will depends.)
My idea currently uses audios cues (and visual cues) to generate the soundtrack and the gameplay is based on it. The concept allows a difficulty selection system to be implemented only with a few lines of codes. I already have looked and downloaded a bunch of images online for references as my idea of the visual style follow a specific kind of look. (It might be for naught if the Jam's theme is too much specific or restrictive, but there are barely any chances it happens.)
As I find my idea really good, I do plan on following up, regardless of the results in the jam, with a commercial mobile release of it with more features so if it wins, it's only adds more onto my project for the next 1-2 months. That's the kind of possible follow-up you can find with these kind of Dev Jams contests. :)
