This is provocative and fun but doesn't hold true for me personally.

If I'm working on only one project I will often end up getting blocked where banging my head against the problem doesn't lead to progress. The only reliable way to progress in those situations is to leave the problem for hours or days and return. Having multiple projects simultaneously allows me to spend that unblocking time making progress on other projects.

I learned that as a lab rat, learning electron microscopy. When sample processing takes months, you need to be working on many projects.

True for me as well.

Too much context switching in a day has costs. But, not switching back to a task soon enough also has costs...you might forget certain details of the current state of the project after being away for a week or two...and it takes time to refamiliarize.

So I prefer 2-3 projects, spread evenly through the week, max two in the same day.