I'm guessing never. This project was a complete failure and a borderline scam.
He said borderline. That means not necesarilly but it sure feels like it. Absolutely 0 of the donators got any kind of ROI whatsoever.
True but now the primecoin community has a working, albeit buggy and inefficient, gpu miner with source code to serve as a starting point.
That's something at least, but yeah, the donations part was rather disappointing for now, to say the least.
Without the donations the GPU miners wouldn't have come out so soon I guess. The donation has worked more or less as donations should.
As one of the donators, I feel a bit cheated. I had assumed that at this point, the gpu miner would be at least a working beta, not the mess we have on our hands, which really does not work. Perhaps mtrlt will release another update with various fixes he's got, but until then, we've just got an alpha at best - not even a good proof of concept. Perhaps it's due to the naysayers that he released it so early in development - and perhaps that's why he's keeping quiet as well.
Ideally, I'd love to see someone take up the code and develop it into a working miner - even if it's just GPU-accelerated - any bit that will help the cpu miner at doing its job would be appreciated.
Sorry your inventment didn't or hasn't worked out. But as donation it worked out.
As a side note the investors shouldn't have rushed mtrlt to bring out the miner. Among the three factors -- quality, cost, schedule -- you can only control two. People demanded schedule with a limited cost, and they get a low quality product.