Wow, I think I've brainstormed this pretty well, but you're a lap ahead of me here!
My thinking on the engine is that with enough attention to ducting, most of the heat issues could be avoided. Deep, blower powder would probably overcome any ducting setup and choke cooling air to the fan, but I suspect any build like this would have too little power and track to get far in that anyway. The old Fuji motors are awesome for what they are, but there is room for improvement in both weight and power. The iron liners also hurt the cooling (on both liquid and fan motors – was just reading about heat buildup issues on early Storm motors). Custom AL-nikasil jugs on the old 488 crankcase might be the ticket if cost were no object... Overall, I suspect the 550F will be the engine I go to eventually, but I've got a couple Fuji fan motors I'm going to experiment with first. One of the main things I want to try is coatings on the pistons and combustion chambers. My hope is that I'll be able to port and pipe a 488F the same as the 488LC, pushing power into the mid-70s, while keeping the weight advantage of no water pump, coolers, or coolant. Hopefully the coatings would be enough to address the added heat; what I've read about people who've opened up fan motors without addressing heat buildup is that the motor will just overheat and lose power or burn up at some point. Down the road (probably way down), I'd like to experiment with EFI on a fan motor. I'm pretty sure I could adapt 90's Poo EFI run by a Microsquirt, and I think such a setup could at least keep the engine from burning itself down without just having to run rich all the time. Most of the attraction of the fan twin is it's a fairly cheap drop-in way to drop a lot of weight; I'm hoping some outside-the-box thinking will eventually mitigate the heat question, and how much potential my ideas have largely depend on that.
I hadn't given any thought to a single-cylinder engine because of the lack of choices without a difficult custom build. Putting aside the heat buildup question, to match the HP of a 550F, you'd probably be looking at no less than a halved 800. It'd be interesting to see how the weight of such a motor would compare. I'd also wonder what the NVH qualities would be; that's well into thumper territory. It may be close enough that it'd come down to a "pick your poison" proposition between NVH and cooling issues. What intrigues me most with the single-cylinder idea is a build around the AC Blast motor. It's too rich for my blood for the time being, but I'd love to see someone try it; that might be the best practical choice for a single. It's "less than 50lb," but that's probably the bare motor. Also vague is "65+ HP-class performance." I'm guessing it's got at least 60HP, but all that's really saying is it's equivalent to current Poo and Doo fan motors. It uses a balance shaft, which should address most of the NVH concerns. I'm pretty committed to fans for what I'm planning, but if I hit a wall with that, the Blast motor makes the most sense for the next step if/when I can get one cheap. Personally, I can't see building a big single when that would probably cost more and be a lot more work than buying a Blast and selling the rest of the sled in parts. For someone with more know-how and equipment, a cut-in-half twin could be a lot cheaper. Anyway, hats off for creating a lightweight single! Hopefully there's more headroom there; just a guess, but I bet the Star Lite case can handle even more HP.
This, plus some snow on the ground, has me thinking about the next step for my project. I've gotta take the jugs from my old 440F to get bored and honed, and I'm working on getting the quickdrive components to convert it to a belt drive. Even if it doesn't work well in the mountains, it should at least be a fun sled to ride around the house when there's snow.
