No, we should mock people for entertaining the idea and being unable to muster enough knowledge and reason to start being very skeptical. It's basically the same sort of estimation that happens when watching Shark Tank and wondering, "Well, that's a nifty product. What are the margins?" Basic Thermodynamics isn't too far off from accounting. There's a big difference between "works technically, in theory" and "would it work in the real world?" Essentially, you're saying it's okay that people can spout basic science, but still can't apply it in their actual lives in a market economy. Shouldn't that be the lower bar?
I'm saying that the idea is completely compatible with science, and at a cursory inspection seems like a completely viable solution. There's no education problem involved in at least discussing it.
Is coal at ~70% efficiency a better solution than nuclear, solar, wind, hydro, etc.? I'm not sure, but that is sure as hell not snake oil.
If it takes you 30% of the initial energy to get to the heat->electricity phase, then the heat->electricity phase receives 70% of the initial energy. From that point, a 35% efficient transfer means 24.5% efficiency from the initial energy.
Whether that's a commercially viable amount or not, I have no idea. And your last point, I agree with. We should certainly make the coal industry include externalities in their price as much as feasible, making for a (more) fair energy market.