For nanotubes I have sometimes used quite a low value, like 25 Ry, and it seems to work quite well. I've never gone as low as 10, however, so I can't really promise that it works that well.
What I would recommend is that you try to rerun a few of the calculations with a proper mesh cut-off but with the 25 Ry solution as starting guess; it's likely that the new calculation converges very quickly, like in 5-10 steps, since you already have a pretty good density matrix. Then you can compare some key results to the 25 Ry values. And if they are very different, perhaps it doesn't take that long to rerun all points, again with the 25 Ry converged calculation as starting guess.
