Solar isn't nearly good enough yet to sustain America's electrical demands. Individual panels aren't very efficient yet, so you need a metric fuckton of them. They're not cheap either -- partly because they consume rare earth elements to produce. It can cost over $10k to install a system that can power a modestly-sized house. One house. Now scale that up to power a metro area of 1.5 million people.
It also requires absurd amounts of land. And massive infrastructure to monitor each individual panel and gather up all the electricity trickling from each panel into a central station to step up to high voltage to send to the grid. That's miles of electrical lines in a big mesh across multiple acres of land. It's a power station with tens of thousands of little generators. That has to be carefully coordinated and closely monitored. The cost of monitoring and maintenance on that alone would be astronomical.
That doesn't even get into the cost of maintaining potentially tens of thousands of solar panels constantly exposed to the elements. Y'know what wide open spaces have lots of? Wind and debris. Wind likes to pick up and carry debris and throw it around. That scratches the glass (or plastic) panels that protect the delicate solar collector wafers. Scratches reduce how much light can get through to the wafers, which quickly starts to reduce their power output by a noticeable amounts. No matter how good the materials are, that's going to happen, so you either have to have inspectors constantly driving around the entire field visually inspecting each panel or a sophisticated monitoring system to identify poorly-performing panels so you can dispatch workers to repair or replace the scratched surfaces. Then there's just straight up damage -- there's no guarantee a strong enough gust of wind won't just chuck a rock into a panel and smash it (or an airplane won't lose an engine that falls and crushes it).
The alternative -- solar towers with a field of mirrors reflecting sunlight at them to generate heat to produce electricity with conventional turbines -- have more potential since they take up a lot less space, are much cheaper than a 100-acre field of solar panels, and operate on well-established and efficient(ish) principles (we're pretty good now at the [whatever heat source]->steam->turbine->generator process), but aren't without their flaws either. They don't produce tons of power individually so you'll still need a parcel large enough for a facility to house the dozens you'd need to meet the region's energy needs. They share the same environmental vulnerabilities with solar panels -- those mirrors are fragile and lose efficiency as they're scratched, cracked or otherwise damaged. Plus they have to be constantly adjusted to keep the reflection angles right. And they melt birds.
Both solar options also share another major downside -- they only work during the day, and only work at their peak output in clear weather. Cloudy days drastically reduce solar panel output and completely stop solar towers (you need direct sunlight for that to work and clouds block it). Storms reduce solar panel output even further. Not to mention the potential damage storms can cause to expensive, delicate components that must stay outdoors and exposed for them to even work at all.
Using batteries to bank power overnight is unworkable at the required scale. The tech just sucks overall right now (expensive, not enough capacity, complex equipment for managing charge/discharge/conditioning cycles, material availability, fire/explosion risks, etc.), and we need some kind of major advancement in battery chemistry somehow to make it feasible.
The molten salt thermal plant you mention is pretty slick (storing heat for energy production during low solar output periods), and they've gotten it working in the field on a moderate scale, but for a facility big enough to power an entire region you'll need a very large-scale system and it'll have one hell of a whopper of a single point of failure -- big-ass tanks of molten salt. Leaks won't be too bad environmentally, but human flesh and bones melt quickly at those temperatures (so there's no direct intervention possible) and it's not exactly easy to replace lost molten salt. It's not just a matter of cleaning out the local supermarkets' stash of Morton's. You'd have to build any such facility into separate units so one thermal system eating shit won't drag the whole plant down with it. Of course, doing that means more expense and a modest reduction in overall efficiency.
Wind power works at all hours, but only if there's wind. The turbines are also godawful eye sores, expensive as hell to purchase, install and maintain, and have spectacularly destructive "failure modes." They're also not all that efficient yet, so again you need massive fields of the things to produce a meaningful amount of electricity with them. You're back to the solar panels scaling problem -- miles of electrical lines and signaling cables for monitoring and a constant need for expensive staff on-hand to conduct checks and make repairs as needed. They also fuck with birds, and attract tinfoil types who insist they hear magic sounds from the turbines that cause everything from impotence to cancer. They're full of shit, but they're litigious and annoying. That adds to the cost of operations.
Hydro's well established and very efficient, but are geographically restricted -- they can only be built where there's a water source with adequate flow to sustain the turbines, plus by their nature they have potentially major implications both upstream and downstream (you're creating a lake and creating a major flood risk downstream if something ever goes wrong). Geothermal is similarly limited in that there's only so many reliable geothermal vents out there to use in the first place. Tidal power is still years away from being practical for commercial energy production.
When you roll up all the up-front costs and maintenance costs involved in all these renewable energy sources, they end up being far more expensive than just digging up coal or natural gas to burn or nuclear material to put into reactors. If it weren't for litigious environmentalists around the world intentionally gumming up the works, nuclear power would be -- overwhelmingly -- the cheapest and most efficient way we have right now to generate electricity. But NIMBY and environmentalist types have made it absurdly costly and time-consuming to get rolling, so we've reverted back mostly to coal & natural gas. Good job, jackasses.