However, if the goal is to generate a 10.000 MHz clock from a 100 MHz clock, wouldn't a 74196 or other fast counter do it "better," meaning less effort, less cost, less complexity, and less risk? If the ultimate goal is simply to have a 10.000 MHz clock, a crystal oscillator wouldn't cost much either.

That was the thrust of my "nonsense" comment.

Clearly, there's considerably more involved than has been discussed. You surely are familiar enough with the issues associated with updating displays, managing data flow, etc, that you can appreciate that it makes no sense at all to use a microcontroller just to create a clock of a specified frequency from a multiple of that frequency, particularly when it really does consume the bulk of the MCU's processing bandwidth.

RE