Hey Patchen,
The question definitely makes sense! This is a very common thing to want to do for a lot of modulation effects (flangers, choruses, etc). Another way of putting it could be that you want the phaser to spend an equal amount of time moving across each octave of the frequency range.
You're definitely on the right track with the audiotaper module. Another thing to consider is which module you're using for the filtering. If you haven't seen it already then I'd recommend using the phaseshift2 module. It's similar to the phaseshift module but it lets you specify frequency values instead of directly specifying the coefficient of the filter. If you use audiotaper and the phaseshift2 module then you should be able move the phaser equally across the octaves by setting the alpha value of audiotaper to the result of depth/polefreq of the phaseshift2 module.
For example if I had the polefreq set to 20.0Hz and the depth set to 2kHz on the phaseshift2 module I would set the alpha of audiotaper to 100. Connect a unipolar traingle wave to the input of the audiotaper and connect the output of audiotaper to the mod input of the phaseshift2 module and you should be good to go. Be sure to set the number of poles in the phaseshift2 module to a higher even number (6, 8, or 10) for a nice phasey effect. Or you can play around with having multiple phaseshift2 modules with different ranges. Each phaseshift2 module you use will need its own audtiotaper corresponding to each polefreq and depth combination.
Larger values of alpha will make it spend even more time on the low octaves and smaller values will make it spend less time at the low octaves.
Hope that helps.
~Woody
