I think we’re actually in (vehement) agreement on the flaws of specific “testing”, whether that’s a 1 dimensional “FTP” test or a four dimensional “4dp” test.
Where I’ve been coming from is that the 4dp at least attempts to consider the four basic physiological systems and provide specificity in targets for each. This is far more useful than bell curve.
I personally would take my 20min, 5min, 1min and 5s power data from an up-to-date PDC (in something like WKO or Golden Cheetah), plug these into the SF app and use that app as my training interface.
I’ve been banging on to the WKO team for some time now that while it’s science is excellent and industry-leading, what the software painfully lacks is a training interface that makes use of iLevels and the recommended interval target/length that changes on an almost daily basis as the model gets more data.