I'm not questioning the findings. I'm questioning the article, and your interpretation to arrive at such a summary.

APL plans to continue to partner with organizations to refine the CHESS thermoelectric materials with a focus on boosting efficiency to approach that of conventional mechanical systems

energy-harvesting technologies for applications ranging from computers to spacecraft

70% improvement in efficiency in a fully integrated refrigeration system.

It's all potential, and possibilities, and future projections. I'm sure someone will find real world applications for it, but a fridge tacked out with Peltier tiles that draws energy from its ambient environment (while actively ruining the thermal gradient by the way) is ludicrous.