1) Virtue signaling. (Not a good reason) 2) The person who builds it truly wants solar but doesn’t own that other land so it’s simple to just use their roof. (Not a good reason) 3) Power transmission from another location more than offsets the cost savings. (I doubt this is true) 4) Tax/policy distortions. (Not a good reason)
Am I missing something? Is there a good reason for rooftop solar?
It's certainly cheaper than buying some land exclusively for it, building supporting infrastructure, and then transmitting the power to the place you live.
Retrofitting a roof with solar panels is a different situation, but I won't be surprised if the costs of dealing with all of the power generation, transmission and distribution companies add up to more than just putting some panels there.
It is an action that can be done by an individual that doesn't require coordination with tens of millions of other people.
This is a real value that should not be underestimated.
With respect to you other points, yes, economies of scale and more ideal sites provide vast advantages commercial solar. Commercial solar farms are ~$1/watt and residential is $2-3/watt with subsidies.
Keeping a grid up and running is a delicate balance between varying consumption and production, like having a tight rope with various needles suspended by the their eye, some with added weight, some with a hand to push the rope 50/60Hz. We normally "average the load" of a grid making large enough but not too large grids with few big generators and many small (or few large but constant) loads. With renewables and a gazillion of grid-tied inverters who tend to be (far) quicker than big ones the frequency keep a bit schizophrenic: inverters see constantly a bit too high or too low frequency and keep adjusting their output generated instability. If they are few others local consumers around them just absorb their output, but if they are MANY keeping the grid up is a nightmare.
I had solar installed last year in Australia. The 6.5kw grid tied system has an estimated payback time of approx six years, even with Australia's appalling buy:sales ratio for electricity (3:1 last I looked). For various reasons, I was not eligible for subsidies, and it was still worth it. I pay around AUD$350 per year for electricity, which is mostly in winter.
In summer, pulling an average of 30 kwh per day off the panel and running that straight into the AC without needing to worry about the electricity bill is amazing.
It's worth noting that due the original subsidies to get the industry going, Australia has some of the cheapest residential solar installation prices in the world. A 6.6kw system (pre means tested subsidies) is around $8,000. That's the panels, the inverter, all the other bits and pieces and the installation. Electricity prices range from $0.20 kwh to $0.40 / kwh.
Another (short term) practical reason is that remote areas where the land is cheap usually have a weaker connection to the electrical grid. As erecting vast solar parks takes months while fortifying the grid takes years (and is very expensive), this creates bottlenecks. Erecting new solar parks in such remote areas is therefore (temporarily) prohibited: https://www.sunforson.com/dutch-provinces-of-friesland-and-g...
Transmission and distribution aren't trivial. Under the right circumstances, people can get very attractive ROI periods, not to mention a degree of independence.
I can't be the only one who learned from the Texas incident that our grids our metastable and one bad crash away from a multi-week outage. I can't be the only one who thought "F that, I want a plan B."
If customers need transmission at night, or even once a year, they still need all the infrastructure in place.
