now, I'm waiting for the FD 15/2.8 fisheye and the FD 135/2 with great expectations
Great results! I agree that manual focusing on the mirrorless systems is a far more pleasant and accurate process, but that is also due to the MF lenses being designed exactly for that purpose. Focus peaking is also a great help but I've found it can be mislead for instance if you have a low contrast foreground and sharply defined background; the focus peaking will draw on the background, indicating it is sharp while it is not really in focus. Best is to magnify and check. So it's a great system for static/still life photography, but other than developing a good pre-focusing technique, it's not that great for dynamic subjects.
The speed booster really is magical; an aps-c camera suddenly really gives that FF feel and it has completely changed the way I look at my NEX-6.
Regarding wide apertures: I've found that wide open, contrast is reduced and there is some ghosting with the FDn 50mm 1.4, but when I checked that lens with a regular adapter (no glass) it turns out to be caused by the lens, not the speed booster.
Keep me posted regarding the results with wide angle lenses. The speed booster performs really nicely with my FDn 28mm f/2.8, but the Sigma 21-35 results were pretty much unusable - worse than without the speed booster. This is likely caused by the lens though as I have rebuilt that one due to a sticky aperture and I may have done a poor job reassembling and/or cleaning it. I don't really have a way to calibrate that lens anyhow so it's trial and error if I try to adjust that lens and get better contrast and sharpness. Still the results without the speed booster are better with this lens so I wonder how the speed booster plays with wide angle (wider than 28mm).