I've read somewhere about an idea of implementing all Process Builders that are executing on a single object within a single Process Builder, instead of using multiple ones (IMO this is a similar approach to what is generally done with triggers).
Is this approach possible to maintain over a longer period of time? Have you seen it used successfully? Also, is it better from a certain technical perspective, for example, is it faster?
Stop
orEvaluate the next criteria
when a PB finishes a specific branch, so it is possible to enter more than one branch in a given transaction. That's the reason I thought that now we could handle more complex logic within a single PB instead of using multiple. And I agree with you, I'm also in a camp of not using PBs at all, but we already have a ton of automation spread out in them, which gets pretty painful when something goes wrong.