I think the EROEI of the whole process revolves around how efficient your retort is. How much of that heat can you scavenge from the byproduct stream and flue... if you can keep most of the heat in one place, and move the shale through it, maybe you could get the energy return up into positive territory.

That, and how cheaply can you build your processing plant?

The key to knowing if it's practical would be to know that he's running the whole operation on energy he has extracted and isn't using vast quantities of water, natural gas, electricity, etc.

How about shipping it to iceland, or some such place. They could "export" geo heat!