the data seems to suggest that stages have certain inconsistencies (spikes) which could be a big deal depending on what you monitor. There's also, as I understand, no way to test and change the slope which is the kind of thing one would do with an SRM (and presumably other PMs) to maintain the consistency of the data.