I worked the problem a different way, first of all I evaluated both ends of both spectrum (2.2-5.2 trillion for adding cars to get the number of solar farms needed and 5-10 acres per MW rating, this is how I built my range). I believe I have an error in the number of solar farms needed (2-2.5 million farms in my original post), but I have not been able to replicate my math that got me the error. I made this post in sections and at some point realized that 27 MW doesn't make 648 MWh, but I might have missed switching it out somewhere to get the math I got.
Rerunning the math I took the amount produced and needed (~17 trillion Wh) and divided it by the production for one 27 MW site (150 MWh) to get the number of plants and then multiplied that by 27x10.
17x10^12 / 150x10^6 x 27 x 10 = 30,600,000 or 30.6 million acres.
All that aside we are still talking about 75x more land usage before we talk about time zones, day night cycles, distribution of the panels, etc. The big counterpoint that people seem to have is batteries, but we already use batteries and the amount more we would need to provide 24 hour coverage with just solar would be astonishing.
Market forces push business decisions, the only way solar power would be cheaper for the consumer is if it was also cheaper for the business. If solar was realistically cheaper then power production facilities then corporations would be switching to it and probably not drop our end costs because that would just be extra profit. Whether it's a lack of battery capability, unattainable capital costs, lack of reliability, or something else at play, solar power would not be cheaper for the end user or else corporations would be switching to it.
EDIT: Good work on your math.