Given a choice between making $400k a year
Read it again, that $400k is just the first year. Then $800k, then $1.6m, then $3.2m... Until he's made his billions of dollars.
If you are capable of making the first plant (as Rossi did), you are capable of earning enough income from it to make the second, and so on, until you've made enough money to sate your appetite and decide to just give the design away. But it's better than that, because if you are actually selling electricity, not magic beans, it makes your magic beans a hell of a lot more credible. [For example, Rossi claimed (early on) that he was using his device to heat his "factory", precisely to give himself that kind of credibility.]
Can't get on the grid due to {conspiracy}. No problem. You sell on-site off-grid power to individual large customers, if you can cut their power costs they'd jump at the chance. Aluminium smelters spend a fortune on electricity, mining companies spend a fortune on fuel for remote power generation... (That said, you might pick one or two clients as demonstrators and offer them free power. That also lets you work out the bugs in the system before you have contractual obligations.) A decade later, you're powering everything from Google server farms to the International Space Station. And at that point, if you are not wealthy enough to get a licence to build grid-connected plants in any market that has commercial power, you're doing it wrong.
But there's even more money to be made. In the paper, the isotopic composition of the post-experiment "ash" from his ecats included several isotopes that are extremely valuable in their own right ($20k/oz.) There's one that is a beta-emitter that (if it could be produced cheaply) would power compact "nuclear batteries" for anything from laptops to space probes, for applications too small for the ecat-based system. [At least one lithium isotope is "dual use", so selling that requires some extra paper-work. But others are no harder than selling smoke-detectors.]
Right now he could not only be selling power, but also be selling rare isotopes, and developing other product lines not directly connected with the ecats made from the "ash" of his power plants. (And the good thing about the beta-voltaic battery is that not only are the fairly simple, the technology is off-patent. Doesn't matter if someone reverse-engineers your design, unless they have a source of cheap beta-emitter. So you can sell the "batteries" wherever you can get appropriate licensing, without worrying about IP theft.)
And every one of these things does more to demonstrate the reality of his device than getting a few gullible patsies to write stare at a glowing rod for 30 hours.