After talking with several observers, and submitting three nights of BSM_South data, I'd like to remind observers that saturating 5th magnitude variables with regular telescopes is REALLY easy to do. There are lots of V-band measures for Nova Del 2013 that are ~0.2mag fainter than they should be, and lots of others with way too much scatter for a 5th magnitude star. Look at the observations from observer HQA and you will see what I mean. This star really doesn't change on a minute time scale; if you are seeing 0.1-0.2mag scatter, it is all due to the way you are collecting data.
With BSM, a 6cm telescope, saturation sets in at V=6.5 for a 30second exposure. You can scale to your own situation. With a 16-inch (40cm, 44x more collecting area) telescope, for example, you will saturate in less than one second for V=6.5, and for Nova Del 2013 at V=5, the exposure will be in the 0.1sec range before saturation. Don't assume that 65535 is the saturation level for your sensor; many SBIG cameras, for example, start saturating at 45K ADU.
I'd keep the peak ADU in your image to be half-well (30-40K ADU). If your field of view is small and you don't have good comp stars, then stack as many short-exposure images as needed to build up the signal/noise. Stacking doesn't hurt in any case, because anything less than about 10 seconds will also suffer from scintillation, but stacking effectively increases the exposure time while retaining the dynamic range.
We want quality data - the pros are waiting until the nova fades, and this is a chance to show what the amateur/small telescope community can provide!