Your output might be 20 kHz audio, but in the real world you're surrounded by analog at much higher frequencies. I've worked on enough "low bandwidth" audio projects plagued by problems at higher frequencies that I can't really recommend the "Its good enough for audio" approach anymore. You can easily have a tiny DSP with a noisy clock signal in the tens of MHz. Or a wonky USB signal. Or an intermodulation product from some unrelated circuit.
If I can pay $100 for a super miniaturized audio only scope, or the same price for a bulkier 100 MHz scope, I'd go with the 100 MHz scope every time, so in that sense you're right.
That being said, it sure is nice that today's scopes are more like a lunchbox than a boat anchor. My lab at work easily has a million dollars worth of "full size" test and measurement equipment from Agilent and Tektronix, but there's no reason the entire budget has to go to them. We're happy to pay for stuff like this (http://www.triarchytech.com/) when it's a solid product that makes our lives easier.
I'm thinking strictly of things like mixers and junction boxes. I take your point about DSPs and other system clocks (and am trying to figure out if this is an issue on a hybrid DSP/FPGA/discrete device right now), but in areas like live sound engineering your main enemies are things like ground loops, crosstalk between mixer channels and whatnot. These are installation issues rather than product development ones.
Seems like a portable, precision FFT analyzer would be more useful. The ability to measure frequency response, harmonics, cancellation, distortion, noise, etc. make an engineer's job so much easier. Frequency information is far more important than timing information.
What benefit does an oscilloscope have over meter+ears when resolving ground loop or crosstalk issues? Time domain information doesn't seem terribly useful.
Knowing whether a hum is 60Hz or not tells you whether it's a ground loop or some other problem. For crosstalk a lot of mixers have a 1khz sine tone generator. It's useful to be able to verify that you're dealing with a particular frequency. You are not always able to rely on your ears to the extent that you'd like when you're in the field, since the environment where a PA is being set up can be noisy as well.
This device has a few other tricks up its sleeve as well as a basic 'scope; it also does FFT, phase metering, and operates as a voltmeter. I'm not suggesting it's the ultimate device of its kind, just pointing out that there is a class of people for whom having this functionality in a highly portable package would in fact be handy.
Oh, and the protocol sniffer means you can test MIDI output from a synth or sequencer.
If I can pay $100 for a super miniaturized audio only scope, or the same price for a bulkier 100 MHz scope, I'd go with the 100 MHz scope every time, so in that sense you're right.
That being said, it sure is nice that today's scopes are more like a lunchbox than a boat anchor. My lab at work easily has a million dollars worth of "full size" test and measurement equipment from Agilent and Tektronix, but there's no reason the entire budget has to go to them. We're happy to pay for stuff like this (http://www.triarchytech.com/) when it's a solid product that makes our lives easier.