Five years of cards in one app: what to do about the day it stops being maintained

Depending on one piece of software for years of study is a real risk and mostly a manageable one, provided you know which half is portable.

The cards are portable, the scheduling is not
Your notes and cards export cleanly to a plain text or spreadsheet format from almost any of these, and that survives anything. What does not travel is the review history and the intervals, which is the part representing years of work. Different apps use different algorithms, so even where a history can be imported, it is reinterpreted rather than preserved.

What I actually do

• A monthly export in the most open format offered, kept with the rest of my backups. Not the app's own backup file, which is only useful to that app.
• A full copy of the app's native file too, since if it does still work, that is the only thing that restores intervals.
• One restore test, once. Everybody has backups; almost nobody has tried them. Mine failed the first time because I had exported one deck rather than all of them.
• Keep media in the cards rather than as links to something external, or you lose half the content silently.

What actually threatens you is not the project disappearing, since mature tools with open formats tend to be forkable and often are. It is a platform update breaking the mobile client, or a sync service closing. Both are survivable if you can rebuild from an export, which is the only thing worth checking.

If you are choosing now: pick on whether the data leaves cleanly, not on the scheduling algorithm.