Ok - I'm going to go along with your argument for a little bit here.
In the Texas example, they were referring to extremely high prices for 1 hour in a year. It was a record. What you are saying is that this makes up for the other 364 days in the year or the other 8759 hours in the year.
That's not what I'm saying. What I'm saying is that the energy provided by solar panels in areas with high loads during the afternoon is generally displacing more expensive energy. The only exception would be if a consumer used all their energy at peak demand, but I know of no one who uses electricity like that.
In CA, and probably most other areas, the cost of a grid connection is very low, something like 1c/kWh or less based on the data in CA from my last post. What this implies is that as long as the value of the energy generated by solar panels is ~1c (or less) more than the value of the energy used by the customer, net energy metering costs the utility nothing.
Or maybe more realistically, in CA the utility pays $1 an hour for 3 hours a day for 14 days a year. So the 42 hours of extremely high priced energy makes up for the 8718 hours that aren't? You are at .5% and you are talking a 25 times increase in price - that is a blip. The vast majority of the time the solar is displacing 4 cents an hour electricity. So sure, it helps that solar is generating well during those peak times but making the average 5 cents is probably hard to do even with the 2 weeks of heat wave.
It's not only that the utility pays $1/hour for 3 hours a day/14 days a year on the high end, but that they also pay something like ~$.05-.10c/kWh for load following natural gas in the winter on the low end. This compares with baseload at only a few cents per kWh.
Because most consumers use energy though out the day, with some of that being more expensive on-peak or load following, and some of that being less expensive off-peak or baseload, in the aggregate, this implies that the levelized value of the energy a consumer generates is going to be greater than the cost of the energy they use in regions like CA. As long as the difference is ~1c/kWh, there is no subsidy for the consumer.
In cold regions where solar panels are only competing with baseload, and the utility has night time peaks, then net energy would be a subsidy. But in most places where there a higher A/C (and other) loads in the afternoons, it probably isn't given the current level of grid penetration. In fact, I would imagine that most net energy agreements represent a cash benefit for utilities because the difference in value between energy used and generated in most places where we have a lot of homeowner generated solar power is greater than ~$.01/kWh.
Texas has a very peculiar regulatory market also and they have seen electric rates through the roof over the last few years. For better or worse.
ERCOTT is kind of isolated, and it looks like the average real-time prices are driven by extreme weather events.
http://www.potomaceconomics.com/uploads/ercot_documents/2011_ERCOT_SOM_REPORT.pdf
http://www.ferc.gov/market-oversight/mkt-electric/texas.asp
~$50-$60/MWh seems to be the average, although they have been as low as $30/MWh. Once Tres Amigas is completed, prices for the west, east, and ercott should drop since Texas could tap the east/west for energy if they need it and also build/export a lot more wind power.
The right strategy in CA is to shut down all industrial use of electricity for those 42 hours a year. Something I am sure they already do. The other easy thing is residential peak demand throttling which they do in NC but only pay people $25 a year to do it (so I'm sure they don't get many takers).
The best strategy is doing both. The problem with throttling demand is that if people/businesses start opting out, then a utility needs to cut demand, and they'll do so in a controlled or uncontrolled manner, which usually comes via normal/rolling blackouts. Both additional generating capacity and DSM are good for the grid.
What always surprises me is that I live in the South. A/C represents less than 10% of my electric needs. We pay a higher peak charge in the summer ($5 vs $3.80) and surprisingly I am running higher peaks in the winter. List of things that use more energy (both peak and aggregate) - hot water, heat pump, dryer, car.
I can see that, although I imagine that there are quite a few people who use gas and have higher afternoon peaks in the summer.
http://www.duke-energy.com/news/releases/2012072702.asp