The system has a 20 year warranty and is expected to last 25 years.
Regardless, technology will be much better by then.
Also, PGE rates will be off the chart; heck they are now.
Perhaps some of the fine calculations others are making assume static electricity costs....
Sunny CA is an optimal place for solar; I believe all new housing requires solar, not sure.
You see panels everywhere.
I am very happy with ours.
Going forward, technology will improve and costs will go down.
This will enable other areas to come on line, if they should choose to do so.
I don't assume costs to be static, but BNGS, which is our largest single source of power (30%) is currently paid $0.067/kWh to rise to a whopping $0.077/kWh once their 6-unit refurbishment is complete. That will provide power until the mid 2060's. Another 25% of our power, which is hydro-electric, is currently paid $0.047/kWh, and not expected to rise significantly. 14% (3,000MW) of cheap ($0.07/kWh) nuclear will be dropping off the grid starting in 2024 as Pickering retires unless something is done. The other OPG nuclear asset, Darlington (16%, 3.512MW) is currently in the process of bringing its first refurbished unit online. It is paid $0.07/kWh and they are petitioning for some rate increases but probably won't get them. If anything, once the insane wind and solar contracts start expiring, our rates may actually go
DOWN 
One of the biggest components of our electricity, as noted in my comparison, is the cost of delivery, which has a set base rate and then an additional charge based on kWh added in. This is directly tied to population density of the area you live in. Rural ratepayers are hit much harder by this. Much of this cost is due to the largest distributor in the province, Hydro ONE, having to fund the connection of privately owned VRE assets to the grid, which, given the size of Ontario, and their distributed nature, was extremely expensive. While both rates and delivery are regulated by the province, the province of course allows for recuperation of legitimate expenses, which this was.
Physical location and market structure have a big impact on the viability of grid-tie systems. Cost is of course a significant factor and, due to superior output profiles at lower latitudes, this is certainly made an easier sell in places like California, Texas...etc.