The topic suggests numerous variants of confounded study design and bias.

- Solo founders may face a steeper selection curve, therefor more bad ideas are weeded out.

- Conversely: multiple founders might get an easier pass.

- Investors might jam together multiple founders who are poorly suited.

- Solo start-ups face lower costs, and have a lower minimum altitude.

- Coordination problems. With only one person, you've got fewer disagreements over what to do / how to do it.

That's based on reading the article, but not the research itself.

(Thanks to dang for the title rewrite, the article's headline is torture.)