Thank you Paul, Olivier. Of course, many thanks go to Steve and François (who can handle the pro-am link even while on holiday!).
Woody, your fit is certainly valid (just a small excess around 4000A). The starlight from HD207673 seems to me very reddened for an A2 class star (see the flattened continuum and the Na D deep absorption), although the Miles catalog report E(B-V)=0. Anyway, if the library profile is the observed spectrum (like any other one in the ISIS library), your calibration is correct. You should have confirmation if you calibrate the nova spectrum with another Miles star. If not, we can try to observe and calibrate HD207673 with a response curve calculated with another Miles star, then compare it with the library profile.
The V magnitude during my observation (from ~ 21:00 to 22:30 UT July 4) was stable enough. See:
Unfortunately yesterday I didn't observe the nova and I don't know the brightness trend. AAVSO reports a few measures and I don't understand if the nova lightcurve was on an horizontal segment around July 4 (so you could use it without making large errors). Umberto observed almost at the same time (we coordinate by phone) while your observation is ~ 14 hours early. I don't know...
I hope to observe the nova again if the brightness holds.
Paolo