True. It is only a working approximation to Ebbinghaus. Supermemo and Anki can get on his curve much closer, but the problem is in order to work they repeat material you really already know and this method enables you to skip writing that again, thus focusing only on the unlearned material. In order to achieve that economy, I have a working approximation to the curve, which works on reiteration of unlearned material. Like scissors cutting the hair too long but then cutting again until you get what you want versus electric shears with a set depth which might not be precisely the depth you wanted and only work if a short cut is desired, and the shears won’t work at all unless you have electricity.
It’s harder to see the curve that it is with the Supermemo algorithm, but by the application of the 14 days with iteration, you do cover the same thing, only with less work and less risk of switching on more short-term memory.
In addition to that I have a scientifically untested hypothesis which is divergent from Ebbinghaus, namely that short-term memory and long-term memory are distinct functions, the one happening when we remember consciously and the other when we allow it to function like breathing when we forget about it. This idea is my own conclusion from non-scientific observation of my own case and the case of users.
The number of contented users of GLM who keep going with it into repeated projects year on year, some of whom have done as many if not more than I have and probably will end up doing much more, proves that the method certainly works, it works probably regardless of people’s views about their own learning type as long as they at least enjoy the process of writing.
In short you’ve made a fair comment about the relation of Ebbinghaus. I want to acknowledge the input of what I learned from Ebbinghaus, but you correctly note it doesn’t stop there.
Not enough real science has been done with memory and if anyone wants to fund or test in a real university setting the hypotheses underlying the GLM then I’m happy to be involved.