The example is extreme, but they also point that out in the article. It's still useful to be able to model this, and as they point out, if you use a less extrem

The example is extreme, but they also point that out in the article. It's still useful to be able to model this, and as they point out, if you use a less extreme but still extremely ambitious 5% of the Sahara, the effect becomes negligible. So in theory, solar power could affect the climate by itself, but it practice it won't. But a model that shows that if you change something in this part of the world, it's going to affect rainfall in these other parts of the world, is still very interesting and potentially useful. It could help reduce desertification.