absolutely right and the negative feedback discussion is very enlightening.
from my understanding its easy enough to make an opamp with wide flat frequency response and low noise.
Well, yes and no. Opamps have very wide frequency response, but it's a rising response. To keep it flat, they hafta take some of the output signal (actually, a whole bunch), invert it, and put it back into the input, so that the negative feedback keeps the output linear. It takes a little bit of time for the signal to move through the opamp, so you get a teensy weensy delay; not much, just a little.
But, that delay rolls off a bit of the top end and plays a little havoc with the rise time. No biggie, since the op amp is good out to the megahertz range
BUT, we ain't talking about one opamp here; we're talking about several opamps in one circuit, with negative feedback around each one
and maybe negative feedback around all of them - from the output of the last opamp, back to the input of the first opamp.
In theory, it should all work fine, but theory ain't what it useta be. Not when you're dealing with young designers, fresh outta school, or mega-companies, looking to save a couple of cents on each piece. Huge amounts of negative feedback can dramatically reduce noise, THD, and IM distortion, but all those negative feedback circuits gave rise to a new distortion, TIM (which stands for Transient Inter-Modulation Distortion).
Hope this clears up some of the confusion.