Thank you all so much for the suggestions and solutions!
dtengineering/Tom: I know a linear regulator would probably work in this situation, but it would produce a large amount of heat (a linear regulator is the only kind i know how to build (zener diode and a transistor), and because of the need for a cooling fan or something of the sort, isn't the most attractive solution for me since this module will be in VERY harsh environments (dust/dirt/vibration/possible external high temperatures (may be mounted near the rear axle's differential)). But, thank you for the vote of confidence on that being a viable solution and for your thoughts.
dtengineering: It would be interesting to try and pull it off with a big 1F capacitor or something and a victor speed controller, but I'd be worried the victor wouldn't update quick enough to respond to the rapid changes in loading.
Greg: I like that these are partially encapsulated, and they'll definitely do the trick, so this may be what I end up using for now... but the price is a little prohibitive (since I'd need two).
Kevin: I think this is a winner. I'll have to look at some efficiency curves to figure heat output, but for now I think I'm going to run with these boards and encapsulate them myself (provided the thermals work out).
Thank you all!
-q