"For a society to run on PV on scale" you of course rely on municipal, commercial, and regional "actual tech" solar and wind installations, storage, transmission lines and backup peaker plants to make up the difference. Some storage might involve lithium batteries, or other chemistries, or non-chemical methods such as pumped hydro or other gravitational methods, compressed or liquified air, or synthetic fuels such as hydrogen and ammonia; and commonly a mix of the above.
Floating PV is sited at places where the water is habitually calm; calm water is always flat. Reservoirs and canals are favored sites, as noted in the article. Nobody talks about floating PV on breaking shore waves. It is hard to guess what could be meant by "string output", if anything, unique to floating PV. Solar panel output is the same, terrestrial or floating, and is handled the same way.
We are fortunate that planning for regional renewable power does not depend on people who cannot see how they would deploy it. Others, including other engineers, can and do.
Floating solar will be sited throughout most of the world. Ice on a solar panel does not reduce its output much, but being embedded in ice that, finally, breaks up could be damaging. Engineers are good at designing for local conditions.
The article makes a big deal of imaginary shortfalls of available land for solar, but there is in fact very far from any such shortage. So, solar will always be sited wherever it is cheapest. Some places, it will be floated. Other places, it will share pasturage or crop land. Others, commercial warehouses. In France it will be, by statute, on all big parking lots.