Agree that most notes from other documents don’t make good cards. Making your own cards is worth the time spent, for the most part, although there are plenty of standard decks you might as well clone (Greek letters, as a simple example.)
Anyway, Anki scale in the sense that you don't have to do thousand cards if you don't want to. But the UI was designed to help combat psychological problems of tackling an intimidating amount of cards.
If you plan on sustaining dozens of new cards per day indefinitely, you’ll bury yourself in reviews, but spaced repetition does ensure that if your recall rate is high, review count degrades over time, and eventually becomes negligible. I go through periods of daily new cards when I’m focused on learning something, and periods of review and consolidation. If my daily reviews get over 200, it’s time to let the new cards rest for a while.
There are definitely many things that Anki is _not_ good for. But for the things where it's good, there's nothing else like it. The UI might be stale, but this is outweighed by the add-on ecosystem.
Would you please try the app out...?
1) it doesn't scale. You'll be overwhelmed with repetitions once you're in the high hundreds, and you will be if using Q/A style questions. And what's your daily strategy once it hits thousands?
2) I've converted all my notes from libreoffice writer to anki using python and various formatting tricks, and while the import outcome was perfect, most of the notes are useless - too ambigious/too broad/WIP - you'll never remember those well.
3) premade cards are mostly useless for long term - say you want to remember linux stuffs - you can find cards with thousands of items, but to what point - there's only so much time in a day and you want to learn something different than the author did - you def won't spend a year reviewing them.
I myself am confident this problem can be solved with supermemo and creating tailored q/a from my notes and thinking hard what I actually want to remember long term.