Well, that was fast. I got a nice pretty sine wave output from the board at a selectable frequency. Don't have a scope, but plugged it into some speakers and sure enough it sounds like a clean tone. Of course, I'm not sure it matters for this application. It makes the most sense to pump the same signal to 2 electrodes (the same ones used to detect eye movement) and use a 3rd as a reference, since I don't really have a simple way to supply true alternating current in a regulated matter. I'm pretty sure the brain has a frequency following response at least for a little while (long enough to "wake" the necessary functions),  so true alternating current shouldn't be necessary. Waiting for an extra set of lm317s to come in so I can test to see if current regulation wrecks the signal. I also just ordered the instrumentation amplifiers, but those won't come for almost a month (cheap from china, expensive and still slow from everywhere else).