I think sometimes statements like this probably apply in aggregate, as in you can’t safely say they work in the general case due to too much response variance. But it’s not unrealistic to think it works well for some.

It's not perfect, but it's better than not doing it. He enjoys it.

That is important. SRS doesn't work for many people because, in short, they don't really enjoy it.

When I start to dislike opening up Anki, it’s usually because my cards are too complex and take too much time to memorize. Creating multiple cards with single cloze words makes the process much more enjoyable. Anki has helped me so much with memorizing programming language syntax for example.

Can you share some deck?

Exactly, and that lack of enjoyment comes from the learning content in their SRS app rather than the SRS itself.

I showed my son Anki when I was using it to memorise things for an exam I had to do. He has been using it very successfully to memorise things too.

I'm not certain that the research on forgetting curves has nailed down the timelines well. I know the Supermemo creator thinks his algorithm is best, but I suspect there's a lot of wiggle room and individual differences. Who besides Ebbinghaus has published rigorously on it?