>If only you knew which one!

That one is actually pretty simple: pick the one that you have the most to learn from, because that's the whole point of doing a side project if you are trying to leverage it as something that would help your knowledge/career development (as opposed to doing a side project for some supplementary income, for example).

And yes, I am aware that there are many unknowns, and when you are trying to pick a project with the goal to learn the most, sometimes the projects where you think you already know 90% of the solution end up being the ones where you know the least and have more for you to learn than projects that had initially more unknowns. But without the ability to know the future, picking the one that has the most "stuff to learn" (according to your first assessment) at the initial discovery phase is imo the most optimal strategy. It is fine if you discover later that the other project had a bit more to learn than the one you picked, because the goal is to learn a lot averaged down over a long period of time.

Doesn't matter, just choose one that shows potential and do it. The only thing that matters is that you have an end-product in mind that will provide more value to the users than existing alternatives AND that you actually implement it AND the user finds your product.

Which one do you think about when you're in the shower?

Usually there's one project that you just can't get out of your head. If you don't know which it is, try to take a break from all of them, and you'll quickly learn which one maters.