I agree that GP's take is broadly wrong, but there is a sense in which it does introduce a single point of failure: the sun. If everyone has solar panels, an overcast day zaps an entire city/region.
All this really only shows that the single point of failure fails quite regularly, to varying degrees.
It reminds us that widespread personal solar panel deployment reduces the total amount of centrally generated energy required, but doesn't even make a dent in the max capacity, which is much more important in terms of deciding infrastructure investment.
It's clear from context that "failure" here means a state that is no better than if there were no solar panels at all, i.e., complete dependence on central generation.
