Thanks for your replies. I'd like that nice high voltage differential probe but I can't justify it. I might need to take a measurement like I described maybe once every three years or so (I learned how to troubleshoot 99% of the problems without looking at that waveform).
It seems all the considerations for doing this are the same whether the scope is digital or analog....
Just a comment about the bandwidth / aliasing discussion ... whenever you're working with power inverters, choppers, etc., about the only place you ever see anything that resembles a sine wave is at the output, and that's when everything is working right.
Often what needs to be caught is the amplitude of a steep spike, as across an SCR in a commutating circuit, and often when there are problems to troubleshoot, transients are involved that can destroy devices. There's no way I can be sure there won't be high frequency components on a waveform, generally. That's one of the reasons why I appreciated that discussion. That said, we generally get by on mediocre equipment, helped by a knowledge of its limitations. So thanks!