Kidding aside, flame source heat engines work actually pretty well, as the temperature difference is high enough. TEGs are very inefficient (and easily burned out by overtemperatures), but rocket stove with a heat engine (Carnot cycle or similar) might be viable. Not only do they use very little fuel, but they burn quite clean. Might even be viable for rural areas in developing countries, since you can cook on it and charge your devices.
Of course, that is not flame-source photocell tech.
Sorry, the following is OT.
I was amongst a group shown a demonstration of a wood stove CHP unit that had four "Linear Stirling engine/generator" units and also had a flue boiler for water heating.
The creator was looking for a second industry partner to provide a suitable power management and storage scheme for the electrical side.
The hot side of the 'engines' are in the flame path of the fire box.
The heated water was passed through a traditional HW tank, then two domestic 'radiators', when cooled its routed back to the 'cold' side of the heat engines, before returning to the stoves boiler, for reheat.
The 'secret sauce' of course being the design of the power units.
The creator made no extravagant overunity type claims, saying that the modified nominal 5 to 6kw unit had reliably produced an
average @600W of electricity, and the rest as hot water,
depending on the optimum operating conditions and quality of fuel.
The claim was that running this size of the CHP unit continuously, it could theoretically provide hot water, heating, and electricity aplenty for a small home, in a typical UK winter.
The down side of course is the tending, cleaning and maintenance of the burner, and also a small amount of noise when running.
There is then the problem of having to justify the release of CO² in its operation... not so much if it is used to replace other older, less efficient class 1 heating systems, but I bet even if the creator succeeded in his project the climate change lobby would rightly or wrongly piss all over the idea.