I would have a general purpose LA/DSO and an opportunity to buy a pro USB-protocol analysing SW, would that combination be too inferior to justify buying a dedicated USB-analyzing tool?

The question is faulty. Regardless of what the equipment is the determininig factor is: If ([(debugging hours without device) * cost of hour)] - [(debugging hours with device) * cost of hour)]) is more than the cost of the device the device should be procured.

This is true whatever you discuss a $20 multimeter or a $50k scope.

As an example, most of my ICEs have an optional (against added cost) trace feature. I would lose time every day without the ICE, and once a year I would save a bit of time had I the trace. Thus I do not buy the trace. Other guys I know gain large amounts of productivity from having trace. So they should buy the trace feature, I should not.

What USB tools do I have - NONE. The supplier of the USB chips I use provide all software. I do not have the time to investigate new things for the forseable future. In a few years I may integrate the whole kit and kaboodle and at that time the situation will be different. My comments re USB, if any, are from watching the manufacturers engineer (as far as I know I have made no USB specific comments), my comments re protocol analysis are my own.

Erik

PS do not make the mistake of mistaking salary for "cost of hour" if you are so busy debugging USB that you lose a customer that would provide revenue for years to come the cost could be enormous.