There is a wealth of research supporting the effectiveness of SRS. From Want to recall 92% of everything you learn? This algorithm makes forgetting difficult:
Does SRS really work?
Over two dozen studies by Dr. Price Kerfoot, an Associate Professor of Surgery at Harvard Medical School, conclude that it does. Prof. Gabe Teninbaum, a Suffolk Law faculty member, agrees: “by using spaced repetition, users are projected to remember 92 percent of the material … and if that’s not enough, it ultimately takes less time to study and learn using spaced repetition than other study methods.”
“By using spaced repetition, users are projected to remember 92 percent of the material”
Oddly enough, learning actually seems to be increased during testing, where we’re forced to recall information ourselves, rather than studying, where we simply review or re-read the material. That’s part of why SRS is so effective. It forces our brains to make the connection and try to retrieve the information, rather than just reminding us about it over and over.
Of course, SRS is only part of the process. You must also actively use the vocabulary you learn in context and in the real world, or your brain will not be able to use it naturally in the right situation.
Talking about context, it's worth noting that SRS is just an algorithm, and there are countless ways to utilize it with varying results. For example, you should try making each SRS card a sentence, not just a single word, as well as experiment with cloze tests.