It's not just a 10% increase in productivity, it produces fresh water as a byproduct:
Furthermore, the photovoltaic leaf is capable of synergistically utilising the recovered heat to co-generate additional thermal energy and freshwater simultaneously within the same component, significantly elevating the overall solar utilisation efficiency from 13.2% to over 74.5%, along with over 1.1 L/h/m2 of clean water.
Another commenter summarized the nature article linked in comments...
Yes, the salt is left in the pipes, so they are flushed out at night to prevent buildup.
You're assuming full production for 24 hours a day, I don't think that's likely. Maybe 8 hours of full production a day under optimal conditions? Still, ~200 liters a day of potable water seems quite big for a 5x5 area of solar panels.
It's easier to see the impressiveness of it when you realize that it collects 10% more energy than the current designs on the market. Yeah, that's a huge jump. Typically you only see less than 1-2% jumps in any given technology unless you develop a really novel approach (which is what this seems like).
It doesn’t make sense to think of it in terms of how much of the Sun’s energy it uses because solar energy is essentially free and unlimited, it comes from an outside system, we don’t need to mine it or carry it or anything and we can’t ‘waste’ it in the same way we can other fuels. All it tells us is the maximum theoretical limit.
10% more energy from solar means a rooftop array could generate an extra 300-500W which is a genuinely useful amount of energy.