oh yes, I quite recognise that the time domain response of the 'side' filter section is in no way mollified by the reduction of 'effective severity' (If I may use that phrase rather than either bandwidth or 'Q') of reduced cut or boost. Also that transient smudging and other effects are governed by the filter itself: only their amplitude is affected by cut & boost, whereas in a truly variable -Q- filter section, the time domain response is also altered. By this means, what I refer to as 'constant bandwidth' really has a constant-Q filter, and what I refer to as 'Constant-Q' really has a varying-Q filter. It is the summed result to which I refer and -once again- I'm stumped to come up with a "more consistenly betterer" terminology which can satisfy both the users and the electronics designers/engineers! -Since I deal with a lot more users than circuit designers, I've adopted the habit of using the descriptive terms which address their way of looking at it.
Again, the time-domain issue might be missed by a lot of folks, and I probably didn't even fully appreciate the implications when I first started welding resistors, bits of silicon and capacitors together... If you only look at the overall response graph it's understandably easily overlooked.
In the case of the variable shelf entry slope, the 'Q' control also affects the time domain.
I almost pulled a Capricorn out in 2001 and replaced it with an Oxford. -It should probably have been a very quick swap, since both systems tended to have the same cables run to the same places and the existing MADI runs would have been hard to pull out & replace. Also the full analog & digital patchbays in the control room also had the cabling running to the right locations... -very convenient!
I certainly recognised right away that the Oxford was a significant improvement over the Capricorn, though I -along with several others- was also struck by the nagging feeling that the world wasn't really 'converging' along this operating methodology, and that the place in music recording of the fixed-architecture 'Console-as-hub' scenario really didn't seem to be panning out universally. -Sad, because -as you imply- the excellence which had been developed was also being 'lost'. -Not by necessity but usually the case, none the less.
In the end we didn't buy one. This was right at the end of the official support period from Sony, though the unit itself appeared to be a good deal more stable than the Capricorn which it was being considered to replace.
Keith