Although it shouldn't. Adding salt to water is endothermic, it absorbs energy to add salt and releases it to separate them. The article talks about the trouble of breaking the bonds but this is wrong. It's harder to break the na-cl bond to make salt water in the first place.

We just haven't found a good process. That's it. Cheap desalination is not physically impossible. The deep ocean separation that someone here mentioned seems quite promising and gives off energy which is expected. Salt has weak bonds in water and when they reform the na-cl bond there's a lot of energy to be had.

If solar was cheap enough, it would be used.

In a handful of places, it is. In the rest, it's so much more expensive than alternatives that it's borderline insanity to use it.

Like ice: ice is cheap in Antarctica! But importing Antarctic ice to a desert would be absurd.

Solar is cheap when active, and the good news this is one of the cases where intermittency doesn't really matter. We should really explore more use of intermittent power for cases like this where the limited reliability of wind and solar is less significant.

It's still significant if it means that your desalination plant is only active 50% of the time, since it means you need twice the capacity.

True. This condition only holds in situations where the energy costs dominate the physical plant costs

Fortunately, water usage is correlated with sun intensity.