You raise some good points and you ultimately might be right, but from my perspective, you're recycling lazy arguments without much backing.
I live in New York state (USA) and we have restrictive policies about how much solar we can build and push back into the grid. In some cases, even doing an investigation to see if an area can support pushing a significant amount of solar back onto the grid can cost upwards of $10k (so I've heard). This steers solar production to only provide solar to the facility that it sits next to and disincentivizes pushing solar back onto the grid.
Let's say I wanted to build a solar system, at dirt cheap costs, that could meet 3x the energy needs of my house. What do I do with the excess that can't be pushed back into the grid? You're absolutely right, I could smelt aluminium, scrub C02, or open a "smart" laundry mat, but all those require a large up front capital expenditure.
Put another way, let's say you're producing an excess of cheap energy as an individual, so maybe 100kWh per day. What technology can you install that's cheap to set up, can use the energy in a linear fashion (can scale up) and be profitable? There's other benefits, like transportability and relatively sporadic use, that might come into play as well. I would genuinely like to know what other technology could fill this niche. Cryptocurrency is one and it seems well suited.
In terms of sporadic energy availability, I've heard this argument in a video that Elon Musk did as well and I'm not sure I believe it. If the energy availability is consistent enough, say in a block of 6-10 hours in a day, then I could imagine that could provide enough stability to mine intra-day. Regardless, battery technology is coming down in price and it wouldn't be hard to install a battery system to give consistent power by increasing the solar installation cost by a factor two to three times.
In terms of your basic economic point about Bitcoin using energy that isn't available for anything else, that's like saying energy invested in supporting the networking infrastructure can't be used for anything else. The energy invested in the system is so that the system can be used as intended. If you don't believe Bitcoin/cryptocurrency has any value, then just say that.
In terms of the Visa/scaling/energy/transaction limit, yes, Bitcoin can't use "layer 1" as the payment processor at scale, but maybe providing "layer 2" solutions, like Lightning or something similar can work out. This is what places like El Salvador are using and is why people are invested in Lightning development.
Again, you might ultimately be right, but my plea is for better discussion on this topic here. Especially for the "bitcoin mining needs consistent energy availability", I haven't seen any good resources to really refute or bolster this claim.