Just a general comment on the issue of opamps.
There is a wee bit more to this issue than just having access to a part number or two.
The voltage rails in particular will limit the number of possible opamps that will work in a given circuit. Most anything will work with +-12 Volts and up but when the rails are at 0 Volts and +5 Volts the field narrows drastically. Several of the "hotter" opamps simply will not work with 0 and 5 Volts for lower and upper rail. These voltages are pretty common in electronics of today. The sbm-1 for example has +-7 Volts on the rails. What is possible in this unit just isn't an option in newer Sony units.
The rails also, to some extent, limit the number and effect of known tricks for getting the most from op-amps (Ref Walt Jung's work, POGEE etc).
Then there are issues like source impedance, tweaking the feedback loop, supply rail bypass, shielding, stability problems. Susceptibility to rectify RF energy that WILL leak in.
People talk of picking "fast" op-amps but what use is that if the implementation suffers from huge overshoots and/or inaudible hf oscillations from faulty feedback caps, use of sockets, poor supply bypass and lack of groundplane? Grainyness and hard sounding opamps aren't always because the op-amp itself is inferior in any way...
Ohhh, and current feedback opamps ... :-) Elantec revolutionized the opamp long long ago but it's still not often considered as a general purpose amplifier... The whitepapers (at intersil) are interesting for the technically inclined.
The upshot?
A) There is much more to this than just substituting parts.
B) _Much_ can be done just by picking a reasonable opamp and optimizing around it.
Jon