Yes the spectra of novae change dramatically with time, particularly during the first few hours or days. I have been adding my spectra to the BAA spectroscopy database which show how it has been changing. They show the same trends as you are seeing. See here for a summary plot
The instrument response is difficult to get right because of the high air mass at low elevation. I am now using the nearby star HD29526 as the reference with the A0v Pickles spectrum.
There were Fe emission line curtains in the spectrum but more recently they seem to have disappeared and other lines have appeared. See my posts in the ARAS forum
The velocities in the H and He profiles are too high to be from a rotating accretion disc so the outer spikes in the profiles must be from the ejected material. It is possible that the central spike is from the disc if it survived. (Narrow emission lines like this are a feature of dwarf novae in quiescence) It is also quite possible though that all the features are from the ejected material. Novae explosions show sigificant asymmetry of various tyes and there are various models which could fit what we are seeing. The profile shape is similar to those seen in the novae reported here for example
See here on the ARAS forum for some discussions on possible models
I rather like the last one, the “double smoke ring” modelled by prof Steve Shore!
Edited to correct a blizzard of typos !