You can use 630nm rather than 660nm and it works no problem. Actually it works better because currently 630nm are available on more efficient chips than the 660s. Once we get our hands on the new high efficiency 660s we will experiment to find the most efficient ratio of 630s:660s.

As far as blue/red ratio, it very much affects stretch. So we use more blue during veg and also the first few weeks of flowering if you are trying to control a sativa dom. Red photons are more efficient at driving photosynthesis and they require significantly less power from the emitters per photon so it is beneficial to use as much red as you can get away with while still controlling stretch.

White light can help some strains ripen on time. White LEDs also have a convenient blue peak at 440-450nm. As an earlier poster mentioned, plants use photons from all the visible wavelengths for photosynthesis, including green. I expect that the small amount of these wavelengths contributed by the white LED is used very efficiently.

Plants may go through some period of adjustment when they are suddenly presented with a light of very different wavelengths. Therefore it may be useful to provide some white even during vegging. Since the white LED provides quite a nice peak of blue, it seems possible that a simple combination of white and red could be very effective. Another bonus would be a more even spread of wavelengths reaching each leaf. I will put that theory to the test as soon as I get the new 660s.