Hi Andrew,
Yes I agree, this is an area which definitely needs more consideration. Even Pro-Am campaigns are rarely specified beyond a SNR figure and in most cases I suspect these sort of systematic errors are much greater than the stochastic ones that SNR estimates, and much harder to quantify. These problems are not confined to amateur observations and databases either. For example the continuum of the ELODIE Vega spectrum originally incorporated into ISIS and which has probably used to correct hundreds of spectra in BeSS, is significantly in error.
http://www.spectro-aras.com/forum/viewtopic.php?f=8&t=1369#p6205
and the Indo-US library spectra sometimes suggested as reference stars have a completely synthetic continuum based on Pickles
http://www.spectro-aras.com/forum/viewtopic.php?f=8&t=1582
Cheers
Robin