A hybrid application probably makes more sense at the moment. at least until the Heat Exchanger is validated. Heat xfer efficiency and exchanger robustness run contrary to each other. This application would be in a challenging environment and the failure mode would be scary. A HEx leak would rob compressed air (bypassed into exhaust flow) resulting in partial loss of power, compressor stall/surge, either of which would cause a flame out.
As they mentioned, it's not a new idea. Per their website, they appear to have a respectable, collective technical pedigree. If they do drink the H2 Kool-Aid (or any gaseous fuel app), a little more efficiency could be gained via fuel heating,
Best of luck to them. There's a lot of dead companies on the road they're walking.