AKDan, I just posted on a year-old thread you had. I searched your name due to a reference from another thread. I have a little project going because of a late swarm 2 years ago, when I went looking for a commercial hive heater and was disappointed in what I found. I'm an electronic engineer and decided to design my own warmer with a very tight custom controller, after a lot of research. I lost my original little swarm (size of baseball at time of loss) but not until late March last year, and only after a 2 hour power failure on a sub freezing day. I considered it a success and set out to do further testing this winter. We have had outstanding results and now I'm encouraged. My heater design goes directly inside the hive (Currently Langstroth) and will shortly fit between 2 frames without displacing a frame, although my prototypes thus far, displaced one. I've always placed a frame between my warmer and the cluster, but this last year, we had 2 hives move their cluster to incorporate the warmer directly in the center of the cluster. So it appears the bees have spoken, so now, that is my plan going forward. Currently we sense temp either in the cluster or just above, and apply heat directly to the cluster low in the box. We tested a nuc last fall with a higher setting around 75 F until late November, to extend the brood season to bolster strength, and succeeded in having brood right up until we turned it down in late November, to dormant temp zone. This nuc gained at least one full frame of bees in this time. We left a second nuc set to the 70's all winter long, and she raised brood all winter long! This particular hive had a fairly high mortality rate evidenced by bees on the ground and the bottom board, but found that the nuc had actually grown with the all-winter brooding with the dead being replaced. So while its a small sample, the results are encouraging and we think we may be on to something that could be of help to your region. My control box with no bees, maintained an internal temperature, well in dormant range, until I removed that newer prototype and placed it in my small swarm, to get some in-hive experience with bees. That switch took place just a week before a power failure, on a sub freezing day, took out my remaining small survivors.
I don't know if its enough to keep your boxes above dormant in such cold environments, but with current ratings in the neighborhood of 40 BTU/hr with 12 Watts, it might. Of course insulation and wind break are necessary. It would be possible to scale this design up to what ever it takes to get yours through the winter if you have electricity near your hives.
So don't give up before we've given this a college try... Warmbees!