Yes, it's true the C grading only refers to lead-accuracy, but some people have mentioned it harder to get zero-backlash on c7 screws, because of the relatively large lead-variation compared to C5 or less. For some reason it also seems to hold more true with rolled screws!? This is not my opinion, but something I have read several times here on Cnczone. I have only used mostly C5, so I can't say if this is true or not.
I'm not sure if Zappautomation refers to my post when he says that oversized balls is not the best way to get zero back-lash, but if that is the case; I was not talking about the DIY approach, but about factory made units from THK and NSK and such. Maybe it's not that common any more, but I think they still use this method in production. If someone know better, then please feel free to jump in.
I have several new ball-screw of C5 and C3 quality without backlash, and only 2 of them have double nuts. I don't think double nut is better or worse than anything. Double nut tends to be quite a bit longer, but on the other hand they seem to have a bit better load rating. Just for this thread, and just for fun I compared two of mine:
NSK W2509******* (10mm pitch) (Load rating) Dynamic 11600N Static: 19000N (skewed pitch pre-load)
THK BNFN 2510A-2.5 (10mm pitch) (Load rating) Dynamic: 15800N Static 33000N (double nut, oversized balls pre-load)
There is of course more to it than basic load ratings, but it could be fun and it at least gives an indication.
I don't actually have the THK one, but I have a couple with smaller lead and 5mm pitch.
Ball-screws are a very interesting subject, something could be talked about for hours. I know I could at least...