effectiveness of power filters.

That would be a great question for Paul at PSaudio. He is very good at answering questions.

Distortion with the Ultimate Outlet in place was 3.8. Output distortion in both cases was flickering between .5 and .6. Does this surprise anyone besides me?

This is not surprising. I believe that you are mistaken re. what the PS Audio UO is supposed to do.
The UO is essentially a balun transformer that is supposed to suppress common-mode noise on the AC power line & allegedly not limit current flow (i.e. not restrict dynamics in the program material). The UO is also supposed to provide some relief from AC power line spikes & surges i.e. it is supposed to provide some level of isolation between the input (wall outlet) & output (the duplex outlet). In that UO box there is no AC filtering or AC line power conditioning.
By suppressing the noise & reducing the AC line spikes/surges it helps to improve the audio electronics dynamic range by not limiting the electronics' noise floor due to unclean AC power.
It's your AC regenerator that is supposed to take that distorted 60Hz AC power signal & regenerate it into a perfect looking 60Hz sine wave that will reduce distortion in the playback.
I would rather use a Spectrum Analyzer (rated to be used for AC power signals) to check the common-mode noise floor at the wall outlet & the UO duplex outlet to see if that noise floor dropped. Also, use an o'scope (rated to be used for AC power signals) to see any power line spikes & surges reduced at the UO duplex output.
I believe that you're using the wrong metric to measure the UO output...

you can read more about the PS Audio UO here:
http://www.stereophile.com/powerlineaccessories/460/index.html#8KzJ2HYBgbR84sk7.97   

You'll be better off getting rid of it.

bombaywalla...thank-you for the information.  i get it now.
Thanks
sean

Never had luck with power conditioners.   I've had many in my system, they all change the sound, but in the end, add warts, that I can live without.

The Chang Lightspeed filters do work well (non-current limiting) at noise reduction in certain cases IMO.