We're talking about the grid, and there's plenty of demand when solar is not available -- nights, rain, snow, clouds, long winter nights, etc.
To be brief, I left it out, but my ballpark guesstimate is that, due to the cost of storage, for the main grid, solar for free is still not cheap enough. So, that is an answer to a question in the OP -- how cheap does solar have to be? My guesstimate is $0.00 per megawatt hour is still not cheap enough. In really simple terms, the grid doesn't want solar, even for free because of the need for storage and the fact that solar is unstable, e.g., due to just a sudden summer thunderstorm. The grid is super big on stability -- a small source of instability, and the whole grid of the US NE can go down, and at least once did.
Another, related point is that often the wholesale cost of power on the grid has been ballpark $0.005 per KWh, that is, half a penny per kilowatt hour. So, already, now, on the grid, intermittent power is nearly worthless, even when it is excess power from a rock solidly stable source.
And why might there be excess power? Because at the generators it's not so easy to adjust power levels quickly. So, when there's an excess, just sell it off where can, get half a cent for it, and go on.
Very much what the grid wants from the generators is rock solidly stable power, 24 x 7, no snap, crackle or pop, not even for two seconds, for years at a time. And we have that, now, that is, at least from the generators. Sure, for the lines and poles out to the burbs, during an ice storm, the reliability can be lower.
Solar? Fine, if it is cheap enough AND there is good storage to make it reliable enough AND the storage is not too expensive.
Again, the short answer, for solar, for the grid -- the storage costs too much.
Again, we're talking about the grid. There are other candidate uses for solar power -- taking salt out of water (get to store the clean water), pumping water uphill (if have a big dam handy), getting hydrogen from water (get to store the hydrogen or just pump it into a pipeline), making gasoline from coal and water (get to store the gasoline and pump it into a pipeline). So, these uses all have the feature that there is a good way to store the results of the solar power.
People have worked hard on storage for a long time, and that there is little or none installed on the grid indicates that so far no one has a good approach to storage.
Likely, get the solar panels with 100% efficiency for free, and the issue will remain -- cost of storage. A lot of this is just my opinion, but what's been deployed on the real grid makes me suspect I'm basically correct. We can be sure that lots of smart engineers have checked the figures very carefully, and so have the public utility commissions, so we don't really have to go over all the figures ourselves. It's pretty clear that the smart engineers that have done all the arithmetic carefully don't want solar for the grid.
But, sure, some people want to tax carbon, and tax it enough that solar with available storage will be cheap enough. That's another issue.