Vista,
I haven't used a Prop-1 controller before, but I've used the BASIC Stamp II that the Prop-2 is based off of.
Several years ago, I made a Frankenstein display that had Dr. Frankenstein (2-axis neck, air cylinder for arms raising up together), The Monster (air cylinder to go from lying to sitting up), and Igor (air cylinder in knife switch on the wall). The BASIC Stamp II was connected to an original SSC (8 servo, Serial port control), and a self-built relay board. So, the relays triggered the air cylinders, and the SSC controlled the servos in the neck. Also the SSC was controlling a servo set up to press play on a CD player - I think... although it could've been a minidisc player, cuz it was
that long ago.
Unfortunately, as always, this came together at 5pm or so on Halloween. So, programming time was... umm... limited. The quality of neck animation took a hit, as a result.
I'm not sure how people do this with the Prop-1 nowadays, but when I did it on the BS2, with all the programming done serially, it was kinda hard to make any changes. If I noticed that a part earlier in the show wasn't sync'd up quite right, I'd have to shorten that, and then lengthen the delays afterwards. Also, unless you used something like a timed lookup table, it's probably pretty difficult to animate coordinated motion on multiple servos. [Someone please edumacate me, if I'm wrong]
The other route I've tried is, like hedg12 said, using VSA. IMO, going this route makes this
much easier to animate coordinated routines, and easier to change things, in general. To do things this way, you'll need a computer running VSA, and a servo control board. In this case, you no longer need a sound board, because the computer provides the sound (from .mp3 or .wav). You animate by creating 'events' for each servo, on a timeline. Once you've got your audio, and head animation, you can go one of two ways for the jaw animation. A) you can run the audio for the voice from the computer into a ST board to run the jaw. B) you can animate the jaw yourself using VSA. Option A) will take about 5 hours less per minute of dialog. Option B) will look better

VSA also has a feature to auto-animate to audio, but IMO, it still needs tweaking after the fact.
The only bummer about using VSA is that methods for automatically triggering routines are either expensive, or impractical (again, IMO). Without any additional hardware, you set off a routine by pressing play in the software, or hitting 'Enter'.
Hope this helps,
- Hook