I think you're absolutely right. We are starting to reach the age where a combination of large cooperative non-corporate tech projects and the Internet (that, partially at least, enabled them) are putting us in a place where the actual mortality of project owners matters. The "L" in BDFL is a finite constraint.

I think there's going to be an interesting and complicated churn as several major projects under the BDFL model have their Ds succeed at passing the torch, struggle to pass the torch, struggle to realize the torch needs to be passed, or take the torch and do their best to burn the whole project down so it can't outlive them.