I was just reading about this engine on the Forbes site, http://www.forbes.com/sites/billvisnic/2015/06/13/should-tesla-worry-about-achates-reboot-of-1930s-technology/. It's an opposed piston technology. This means that the pistons meet each other in the cylinder head. Granted, it's a two stroke, and it has two crank shafts, but this doesn't have to be the final word on motors like this one. I'm pretty sure, as in I've already figured out how it can be done but can't afford to build one, that it can be done with one crank and as a four stroke.
The levering is more complicated than most folks are comfortable with when you shed the second crank. The thing to keep in mind, though, that makes it possible to change the Achates design up, is that you don't need as much leverage to make the compression stroke happen as you want to provide for the power stroke to make things go.
Think of it like this, when cars race around a track all day in one direction then there isn't very much call for them to turn the other way. Building a car to do this uses something called weight jacking. Weight jacking leans a race car into a turn ahead of time. It can be done since there will be no call for the car to have to unweight in order to turn the other way. This same concept can be applied in order to lever the power such that more of it is used to make drive power, and less of it is used to compress the pistons together.
Anyway, I bring this up because I think the basic design of the Achates engine seems like it would be interesting to people here. Obviously, I am interested in this for my own reasons, but I'm not asking you to follow my way. Maybe the way Achates is going is the right way. They seem to have a functioning prototype, at least. And it is very interesting how a mix of IC and electric was suggested in the article as a better solution going forward than one over the other.