But if my assumption is true, you might burn a bunch of coal to make these windows, and then never offset that energy use (because the windows break/stop working after X years).
the blinders people put on talking about this stuff are incredible
It would be great if the structure of the incentives could be adjusted to reduce waste and fraud, but accepting that there will be some level of abuse may be the price of encouraging progress.
Fair question. This was precisely the problem with early solar cells. I think one has to take a multi-decade view on some of these things. If the science tells us we can improve the technology within a reasonable time horizon and achieve positive ROI, then it makes sense. Otherwise it is a bad idea.
Side note: We don't want UV light inside buildings. If this kind of glass makes economic sense and it can filter out UV light while producing some electricity, it is a win-win. I believe nearly all glass used in large buildings has UV cut coatings/treatment. If the cost is comparable this could be a viable substitute. The problem, of course, is that this might not be viable as a retrofit but rather something you would include in new designs (you need a new electrical design for the building in order to connect all glass to the grid).
Survival bias. Nobody is thinking about the bulb they haven't had to change in 20 years.
That being said they can't beat regular crystalline silicon solar cells on cost. Their major advantage is that they can be made transparent and lend themselves for usage in multi-junction solar cells.