Yes, somebody years ago published a study that said, Uneconomical! because lift costs, panel cost, transmission inefficiencies.
All of those are now drastically changed, by orders of magnitude. It is not a matter of IF orbital solar is economical, but WHEN.
400W solar panel costs $200 and weighs 40 lbs. If Starship gets down to $100/lb, that is $4000 to put a panel in space. There are thin film panel that weight way less but cost more that would improve the launch cost but still be expensive. That means that will be cheaperp to have many panels on Earth and have enough left for storage.
Every number you mention is going in the right direction to favor orbital.
Don't imagine space-solar will look (or weigh) anything at all like a terrestrial model.
Solar input in orbit is higher than what is received at ground, but similar order of magnitude. I am not sure there are real world designs for beaming power 100km from space, but you are going to take some amount of loss in transforming the power to a transmissible form, beaming it through the atmosphere, and reassembling it on the ground. Unless you have a pin point death beam, the power is going to be transmitted to a relatively large area, requiring a large amount of land to receive it. Why not just build solar panels there?
Seems far easier to overbuild panels on the ground + batteries.
On a sunny day, it's only something like 30% brighter above the atmosphere IIRC. I guess you beat cloud cover (assuming you can deliver it to ground through the clouds efficienly), but I agree it doesn't seem worth it.
If we had laser power beaming it might make sense to put large PV installations around the Earth-Sun L2 point, so the power could be beamed to collectors on Earth's night side. Each beam would serve a series of collectors as the Earth rotated. The market in the Pacific would quickly be saturated, I imagine.
Witness the clause 'at its zenith'. That means, too much atmosphere is in the way the rest of the day to matter.
In space, the sun is always at its zenith, if you look the right direction. Further, you have the opportunity to convert all of it, rather than the 50% that reaches the ground.
And there's more than enough desert that isn't sand dunes. Even the Sahara is only 25% dunes.
A base assumption is often that orbital solar requires a panel in space, but even that might not be the case.