Okay...
So it would certainly seem as if the low compression is the big thing they're trying to work around. Roughly 17:1 is considered the old-school (meaning primitive engines; widest range of acceptable fuels) compression for diesels. So the alcohol and nitro are probably there mostly to ensure that ignition will take place.
Now, something else that came up since I last posted was that I looked at photos of some of these engines, and they're pretty clearly two-stroke. This limits the effective compression one can get, because the classic four strokes needed are paired up in twos and 'sharing' strokes. Anyway, it's worth bearing in mind-- you might have better results shooting for this small size, yes; but not limited yourself to the two-stroke design. It will be more complicated, but the modellers aren't avoiding it because it's too hard to pull off (in my opinion), but rather because of what I said before--- they're willing to trade off anything and everything for small size, and that argues strongly for no moving valves, etc.
Back to our fuels... In any engine, the goal is to get the fuel to vaporize--- liquid fuel plus air does not equal the kind of combustion we're looking for here. In indirect-injected engines, you get the benefit of simpler injectors, because the fuel is vaporizing in a chamber before the cylinders. In direct injection, as the name implies, fuel is injected directly into the cylinders--- so you need a more sophisticated injector (all things being equal) and higher pressure to achieve the atomization of the fuel. This also becomes more and more difficult as the compression increases... So the thinking may have been, we direct-inject in these tiny engines because it's fewer parts, but it limits the compression we can use before our little injector can't overcome the cylinder pressure. Keep in mind, with a traditional diesel, fuel is injected at the top of the compression stroke (or just before the top), against the highest pressure the cylinder can offer, and the combination of fuel and air compresses instantly. The classic 'tick' or knock that you hear in diesel engines is the sound of the injector firing; combustion takes place immediately thereafter. I suppose these little engines can't have very high pressure fuel systems, so the way they're currently designed is to balance the highest compression possible (within the limitations of two-stroke and the small size) against needing a more complex fuel system to create the needed pressure. Meaning, if we took one of these engines as currently conceived and just upped the compression to 14:1 or higher, the first thing we'd notice is that no fuel at all could make it into the chamber. Let me say, though, I know less about the two-stroke diesels, so another possibility is that, since fuel is being injected early in the compression stroke, it can't be any higher or else burning would start before the stroke was at the top, and some of the energy of the burning mixture would actually be opposing our desired direction of rotation. I'm not sure if that's accurate thinking, though, because we started this thought based on the idea that compression was too low for effective combustion--- if higher compression would suddenly start the burn early; I don't know, the two seem to contradict one another. The nature of two-stroke engines carries a lot of penalties, as you see, while providing the benefit of simplicity. It seems to me that what you need is the most efficient production of energy from your fuel, and the most complete extraction of that energy, bar none--- so I think you should be willing to be more complex if there will be vast benefits in that efficiency and extraction.
Looking at the engine you referenced, my suppositions only go so far because it's a two-stroke... but I also noticed that Enya does offer four-stroke engines, including a four-stroke diesel! It makes no more power than the engine you linked (0.8 hp), so I don't infer anything clearly by its existence, except for the fact that: a., miniature valvetrains, camshafts, and injectors are certainly doable, and b., that you will be able to reverse-engineer the already-existing engines out there to get towards what you'd like to have.
As to your question of fuel heating, very well done! Heating the fuel has a very direct relationship to how thin it will be, all the way up to and including vaporizing the fuel with heat alone! Some who run on straight vegetable oil simply preheat the fuel instead of thinning it chemically. The guy I mentioned also feels he gets a benefit by having gasoline present--- better combustion and soforth, so it's not entirely just to make the fuel thinner. But this is definitely a valuable direction to be thinking in. You can make a waste oil burner with nothing more than a blower and a drip-valve for the fuel--- the interesting part is, as soon as the combustion is happening at very high temperatures (and it will be; it's part of the nature of the thing), and if you're preheating the fuel to the point where it is vaporising even before it comes out of the fuel line, then you can burn the darkest, nastiest old-oil sludge (like used motor oil) so cleanly there are effectively no noxious emissions! I think that's incredible. Check out the Association of Backyard Metalcasters--- most of these guys use this type of burner to heat their casting furnaces... temps of 6,000 deg. F are possible, and at those kind of temperatures almost anything will burn, and burn so completely that there's just nothing left to make emissions... just the basics you'd expect, CO, CO2, etc. their thinking is often based on maximum energy density of fuel (because they quickly see how much fuel they have to burn to melt a given amount of metal)... some of them started with propane, because it's clean... but realized that fuel oil, heating oil, diesel oil; these have the highest density of common liquid fuels (over 135,000 BTU's per gallon), so they use less for the same output of heat. They soon realized that at the temperatures they're putting in there, once it's burning well you could pour roofing tar in there and it will still combust very completely. Exciting work, in my opinion.
Finally (for the moment), gas turbines will certainly work if miniaturized, but of course there are limitations. RPMs and heat both tend to be very high in turbines, which is what leads to their short service lives, and demand for sophisticated and expensive materials. There are many metal alloys and soforth that are so expensive turbines are the only place I know of that they're even used. In that application, they are neccesary and/or justified... what's interesting to consider is that some of these would probably have never been developed, or perhaps created in a lab but never ever used, if it weren't for their practical neccesity in gas turbines. Turbines are an exciting subject to a lot of people, so there's quite a significant body of amateur turbine enthusiasts. They do very interesting stuff, and have establised a lot of what can or will work, and what is realistically impossible (for a backyard tinkerer, anyway). For your own effort, there would be a benefit to sticking with what you can do yourself--- see, if it's too complex to build at home (or, let's say, just farming out the machine work), then it's going to be expensive to build, and that will be a major concern as you get closer and closer to a usable product. Consider: an engine the size of a quarter, that extracted 100% of the energy in whatever fuel put into it (which is, of course, a practical impossibility), while producing no emissions and no noise at all... no maintenance of any kind required, ever. Sounds pretty good, huh? Let's say I have them, and can make more... but the price tag is one trillion dollars per copy. It might as well not exist, for all the good it would do--- the cost/benefit analysis would always be saying, in effect, don't bother building it. Low cost brings its own benefit, though... so much of one, in fact, that if I just offered you a ten-horsepower engine, and it had to weigh, let's say, 2,000 pounds (so it's basically a stationary engine), but was more-or-less emission-free and was more efficient (in terms of power extracted from the fuel) than any other engine in history--- and we can produce them for $300 a copy. That's a world-changer right there.
So, in summation, although I earlier advised you to not be afraid to go for more complexity, if the benefit is there; do always keep in mind that the final thing produced has to be producable, and at a cost that is bearable for the benefits it offers. Happily, I think you can achieve that effect without having to think about it too much--- just focus on your experimenting. Take the example of hydrazine; we considered at the start that it's simply too dangerous to work with. Well, that brings with it that it's expensive, and that it can't ever be in the hands of Joe and Jane Average. So even if we had our quarter-sized super-engine I hypothesized earlier, if it would require constant handling and use of hydrazine, it's basically a nonstarter. So if your own experiments are ever precluded by being too expensive or too difficult to carry out yourself, that's a great hint that, even if you overcame the difficulty in the short term (like, I have a friend with a chemical lab, let's say), it will come up again in the long run, and might be enough to kill the whole project. If you can experiment with it at home, then you'll probably stay in territory that will lend itself to humanly-priced production and final products.
I'm glad that this discussion seems to be helping to move your efforts along...
