The explanation given in my earlier message appears to be incorrect and the wpms indicated, while not 100% correct, appear to be very close to the real values. To be on the conservative side until we have a revised FitalyLetris, I have restored the wpm scores to be the ones indicated on the bitmaps.