I think both Frank and Paul have valid points. I will refer my comments to resistive passive, and Amplifiers to Tube type amps.
Almost all actives have sonic signatures and create problems. Only the best minimize these problems. However, Passives also have sonic signatures, but can be of a different nature.
Whether passives or actives are employed, at least two LF filters, and probably one highpass filter need to be dealt with. I think this is handled differently in both topologies.
Passives have problems in that they must not load the source too heavily, yet must present a low output Z to the amp. But this causes some problems. For instance, if a 25k "pot" is used, it may not load the source, but at mid point (worst HF response as measured on a scope), the output Z is many times that of a descent active. Thus the IC, stray, and the input capacitance of the amp must be very low to minimize HF rolloff. If too low an input Z is used in the passive to help the HF, then problems in the bass region could easily develop, especially if the output of the source is capacitor coupled. (Of course the LF problem can also occur with an active preamp, and the output of the preamp to the input of the amp. But one doesn't normally need any lower input Z than 25k.)
Actives can use the same 25k input Z and not load the source nearly as much (HF region) as the output Z of the volume control isn't directly driving the second IC, and thus the capacitance isn't as high, since even the best ICs have substantial amounts of capacitance. Actives, having an output Z many times lower than the passives output Z, with 25k "pot" used, the second HF pole is much higher, thus less HF rolloff.
As mentioned before, the active uses many parts, and most actives do leave a sonic signature, lots of distortion if not designed very very very carefully.
Another point. Since most amps seem to have an abundance of gain, why duplicate this gain with another active stage, and active preamp? Distortion will only be increased and the gain isn't necessary.
As a suggestion, why not outboard the first stage of the amp?
1) An active pre, with unnecessary gain, would be eliminated. Some might say the first stage of the amp is eliminated, but I think there is a difference between those two comments.
2) The volume control would not have to drive a second IC, which is the same scenario as with the active pre discussed above.
3) Hum problems could be reduced as the power tranny wouldn't necessarily be in close proximity. As a side note, the number of ICs would remain the same.
4) Frequency dependent sonic feedback through the power supply would also be eliminated, which is not the case if the first stage remains in the amplifier (all one has to do is examine all the amp schematics one can download on the internet. This is the case against integrateds. However, they reduce the IC requirements by one, which is a plus. IF the integrated uses a separate power supply for the first stage, the hum problem could still pose a problem.)
The only problems I quickly see with my suggestions would be extra cost, any DC present on the output Z of the preamp (if DC out) and possibly dealing with any parasitic oscillations, which should be easily dealt with.
In closing, I would submit a possible solution would be to export the first stage of the amp, with a volume control, as this would give the necessary gain one needs, another useless stage would not be added, hum would be minimized, and frequency dependent sonic feedback would be minimized.
.
