> There's a policy failure somewhere if the cheapness to produce energy this way is not being reflected in the market and subsidy structure. Does anyone have an analysis of wtf is going on?

We are still paying for renewable energy projects agreed in say 2016 for ten years or indeed 2006 for twenty years. The "cheapest to produce" is a fact about new energy not about what it still costs us for things we agreed to in the past.

I don't look at German data but I do stare at UK data a lot and the model isn't so different. In 2015 this country finished "Allocation Round One" of the current mechanism for these subsidies. Offshore wind farms got about £120 per MWh [in 2012 prices] in that auction and by 2016 the first of these wind farms was running. £120 per MWh isn't a good price by today's standards even if it wasn't in 2012 money, but we have to keep paying that price for those wind farms because that's how deals work (unless you're Donald Trump)

This year, in "Allocation Round Seven" offshore wind auctioned for about £65 [still in 2012 prices]. That's still not a great price today, the Solar farms are significantly cheaper, but given that the afore-mentioned Tangerine Palpatine caused gas prices to shoot up this price is actually not uncompetitive. However, that power doesn't actually exist today, the auction was for rights those wind farms will get built over the next year or three and the £65 per MWh power won't be lighting my home until about 2028.