Back to biasing... I just installed a fresh Quad (same tube family, 6550, but different brand) last night and watched the bias during the 20-min warmup:
0 min: 43mv
2 min: 44mv
10 min: 47mv
15 min: 47mv
20 min (ARC's recommended warmup time): 47mv
So in my case it looked like bias changed the most during the first 10 mins, but still only drifted by 10%. Therefore, I think the idea of biasing them at 0 min and then again after recommended warmup time is a fairly good one... Bias for my amp is supposed to be 65mv, FWIW.
On the blue glow... I have never had a power tube that *didn't* glow blue to a varying degree.