I have now experimented with the main mirror adjustment, see below…
[For some reason my original 640×163 image has become vertically squashed, but I think you get the idea…]. Red is to the right.
Anyway – the important things are:
(i) The spectrum gets wider (more height in Y-axis) as it goes down the sensor.
(ii) As well as this, the overall integrated flux gets bigger – an increase of around 50% from the two extremes explored here. I am assuming the final Isis image has removed the background and so the flux is signal, rather than background+signal. (You might expect total flux to increase just because its a bigger area…)
(iii) However, the spectrum is not so good in the larger-flux, lower-in-image version. In particular, it doest work well with the optimal binning option; there is a strong reduction of blue end amplitude. You can recover the blue end by unchecking ‘optimal binning’ (in ISIS) but – a) I don’t think this is good practice (optimal means… optimal I guess?). b) there is some loss of resolution (narrow atmospheric lines are not quite as deep).
Assuming some steady progression of behaviour from the version on the left (spectrum in top half) to that on the right, I think I will be aiming for something like the left hand image on my jpg below… Here its about 1/3 way down.