It appears that the USB/RS-232 converter in the programming cable just can't get along with the USB ports on one particular computer (unfortunately, it's my main workstation, newer than all the others.)
I brought my radio and cable over to a fellow ham's house who had an FT-8800 & the same Kawamall cable. It worked fine w/ his cable attached to his PC. It then worked fine w/ my cable! We then plugged it into one of my notebooks (not previously tested with) and it worked fine there. Actually, on both machines it had an error or two, but the rest of the runs were fine.
It made me wonder if maybe there was a connection problem at the radio, that repeated pluggings fixed. It couldn't be merely a loose/unreliable connection, for the results at home had been consistent every run; it must be something more along the lines of a ground pin that wasn't connecting at all, but there was still enough signal coming through the shell or something that it worked under certain conditions.
Back home, I plugged it in and found that the situation hadn't changed -- it still refused to work on my PC but worked on other computers. Time to verify that it wasn't a software problem!
I used Live CDs (1st the Chirp LiveCD based on Ubuntu 12.04, then a Ubuntu 14.04 LiveCD, after the other misbehaved with my nVidia video card) to ensure that the entire software stack was the same on 4 machines that I was testing on, and that it wasn't a different kernel that was making it work/not. As expected, different versions of CHIRP made no difference.
3 computers worked fine, but the 4th did not. (All of my other cables/radios work fine on it!)
Oddly enough, I *could* get it to work on the 4th computer in one circumstance: Run the LiveCD in a VMWare Virtual Machine, plug the radio into the 2nd USB port on the top of the box (a 3.0 port) *and* ensure that a hub with other things was plugged into the 3rd USB port (a 2.0 port). Is that weird, or what!?
The need to be in a particular what's-plugged-in-where config lent credibility to the notion that it was a voltage issue, but that I was actually able to get it to work in VMWare in one configuration made me wonder if it was more of some weird signal timing problem. It would be interesting to know with more certainty the true details of the problem, but I feel that lack the knowledge, equipment, and/or large amounts of time required for me to determining that....
I think I will ask to borrow the other ham's cable, and try it on my PC. If it works normally, I will probably take a chance and buy another cable. If it doesn't, I will shrug my shoulders and plan on using other machines to program that radio.