[solardev]

The ROI can change a lot depending on your particular circumstances, though. If you live in a not-so-sunny place with minimal electricity use, a 10-15 year ROI isn't unreasonable. But if you live in a sunny place and pay a lot for HVAC or an electric car, the payback can be much shorter (sometimes 5 to 7 years, especially if you self-install). Time-of-use billing can also dramatically change the numbers, depending on your state/electric utility and household usage habits -- sometimes even to the point where you can use a battery to play electricity arbitrage, buying/generating low and selling back high (if your area allows such a scheme).

Electricity prices are expected to keep going up too. But there is also uncertainty in how utilities will renegotiate (or else grandfather in) existing solar ratepayers. California's solar pricing schemes have changed a lot just in the last decade as they went from minimal solar to solar oversaturation (the grid couldn't keep pace with the # of installs).

At the end of the day, for some homeowners, solar is a no-brainer. For others, it doesn't make sense. It totally depends... it isn't an automatic win, but also not something that should be categorically written off. It's worth investigating, especially for new constructions that are heavily electrified (i.e. use electricity for heating/cooling/cooking instead of gas or other fuels).

But agreed that grid-scale makes more sense, combined with other power technologies.

One minor nitpick: It's unlikely the PV modules (solar panels) themselves will fail in 20 years; you'll more likely have to replace your inverter before the cells themselves break. Sometimes those inverters are string inverters in your house (no roof work needed), other times they are microinverters/power optimizers attached to the modules themselves, which would make it more complicated to replace.