here's a thought:
"Processor" is the new "preamp". With 2-channel sources that are delivered in 2.0 analog, no "processing" needs to take place in the preamp. However, unless the source is LP or open-reel, it's probably started life in your system as a digital signal, in which case it needed to get converted to analog... a form of "processing" done in a d/a converter.
I know that you and most everyone reading this thread know this, but the reason I'm stating this is because these days, with digital sources, virtually every signal at some point gets "processed" by necessity. Yes, even simple d/a conversion with linear PCM involves things like oversampling and digital filtering... some d/a converters apply analog filters, some digital, some interpolate beyond 16 bits, some "correct" 16-bit quantization (presumed) errors based on laws of natural wave-forms prior to d/a etc. etc. There's no such thing as the one and only right and perfect way to design d/a conversion... both on the software and hardware side. And so even the most basic of tasks by any CD player can be rightly termed "processing".
I say this because with digital signals, if the d/a conversion stage is incorporated into the pre-amp, then the pre-amp becomes "a processor".
That's really what is meant when high-end music lovers and film lovers talk about audio processors: most are not talking about black boxes that will intentionally alter the source sound and deviate from the original intent by adding reverb or fancy tricks that correspond to gimmicky labels on the front face plate. Rather, we're talking about things like the quality of the d/a conversion, analog output stage, volume attenuator, jitter filtering, resolution enhancement, upsampling etc. Even tasks which are bit-for-bit accurate like decoding Dolby TrueHD, MLP, or DTS-HD MA into multi-channel PCM are generally termed "processing". What does fall into the old-fashioned idea of "processing" in today's processors are things like THX re-equalization and EQ management... and naturally the algorithms used can have a damaging effect on sound quality... but well designed algorithms may in fact improve things given that your room-speaker interaction *also* changes the sound (room-speaker interaction can IMO be reasonably termed as "passive" processing because it's taking place though not by intent). Is it philosophically wrong to use an active means to counteract a known passive effect like an EQ imbalance due to speaker/room limitations? Not necessarily. I liken it to the justification of applying EQ balancing to restore the raw signal from an turntable cartridge so that it has the proper tonal balance (applying an active change to compensate for a known passive effect).
Having said all that... I've found that I prefer the sound of my friend's high-end Lexicon processor when it operates in bypass mode applying no additional DSP to the source beyond the necessary d/a conversion.
Bottom line: a "Processor" is a multi-channel pre-amp with the built-in ability to handle digital signals as well as analog: It's *today's* pre-amp, for all intents and purposes.