Anecdotally, my Tesla solar panel install (via SolarCity) was intentionally made as large as possible when it was installed back at the end of 2017, and at first it exceeded the electrical needs of both house and Model X. Since I was therefore grandfathered into net metering, I did not bother getting any PowerWalls. Even with 0kw net usage, there was still $30-$35 /mo in various fees. Rather than be upset, I realized that if Tesla offered PowerWalls-as-a-Service I would happily pay far more than that for someone to take my excess electricity when I don't need it and return it when I do need it.
That taught me to quit thinking about my electric bill in terms of "paying per KW" and more in terms of "value for the service". What $ value is it worth to me to have APS providing me with that service just to time-shift my own electrical generation? When I went from 1 Tesla to 2 Teslas to 3 Teslas, I realized the auto-expansion to meet my varying needs was highly valuable. Being able to seamlessly add additional cars w/o needing to get more PowerWalls, more panels, etc was very nice, and made getting each additional car an easy thing vs a complex cascade of permitting, fixed-cost purchases, etc.
When I looked at upgrading the main panel and considered the costs to APS to upgrade the transformer in my neighbor's yard which serves our group of houses, and when I was tangentially involved in discussions of a new large-scale solar farm deployment being considered, it was very clear how very expensive fixed infrastructure can be to build and maintain as an ongoing service.
Utility ongoing maintenance / expansion costs should comprise the majority of each residential electric bill, vs per-kW charges, assuming the utility is providing a high-quality service a) w/o repeated service interruptions and b) with proper maintenance / expansion being performed. My understanding is that in California, PG&E was failing at both a) and b), with the increasing rates now having to pay for the backlog of maintenance / expansion which should have been done many years (decades) ago, as well as pay for the liabilities incurred by their failing to do proper maintenance / expansion. Electrical production may trend to very inexpensive with renewables, but grid maintenance / expansion will mean we should never consider "free electricity" to become a thing. (Remember old "nuclear power will make it too cheap to meter" false hype ... even if nuclear had made production nearly free, which it didn't, grid maintenance / expansion is a primary driver of total cost.)