Not to take a dig at you but to set some context for other language learners reading this, 1000 words in a year is awful progress. It would take you 10 years to have just a semi-fluent vocabulary in a simple (for native English speakers) language like Spanish, Italian, or French, and that's not including grammar. This is the main problem with feel-good apps like Duolingo, giving all sorts of positive feedback for what amounts to very little progress.

Anki + a frequency dictionary + 34 new words/day (with 80% retention due to spaced repetition) is easily doable by our brains to get to semi fluency (~10k words) in a year. Or, take a month to cram the first 2000 words and you're well on your way versus other wastes of time.

I agree it was not especially fast progress. I should have said, in one year of casually using FF and a weekly (very good) Italki tutor, I was able to pass a B1 CILS exam and read a children's novel, which regardless of the number of words learned per day felt like a good achievement.

A back of the envelope estimate is that I spent 500 hours in high school Spanish to reach a similar level as I achieved as an adult in ~60 hours of rote vocabulary acquisition and ~60 hours of 1:1 conversation.

Reaching beginner-intermediate fluency in 20 minutes average per day—half of which is spent talking with a native speaker, not on boring flash cards—is easily attainable by someone who is motivated to learn but doesn't need to go at a breakneck pace, or feels that cramming 2,000 flash cards (let alone creating them) in a month sounds like hell. I was pleased it worked for me with as little time investment as I put in.

I am thinking about getting back into studying to see how chatbots have changed the experience, which could probably be used to synthesize more effective recall prompts. (I, Internet rando, argued to the FF team a few years ago that this would be disruptive tech, but they pursued their upmarket $25/session coaching service instead.)

(One thing I like in theory about FF, vs Anki generally, is that it splits the top n wordlist into sets that are "compatible" (fruit, apple, red) rather than "categorical" (red, green, blue). It feels like it helps to build the inter-word connections. However once these are tossed into the grinder of the spaced repetition algorithm, I'm not sure if it makes a difference.)

Where did you get those numbers from? 10k words is generally considered a CEFR B2/C1 level which already is a fluent level.

Semi-fluent, depending on how you describe it, is anywhere between 1k and 5k words. To comfortably communicate with native speakers about simple topics you need 1000-2000 words.