yes the 30-45 seconds off a 5K is exactly correct (for non-elites)
the easiest way to understand this, ignore the 4% math, is to lookup their old PR in vdot tables, then add +2 to the vdot and see what the estimated 5K would be
vdot tables are not linear so they more accurately can guesstimate results
a ~20:00 5K becomes roughly ~19:15, there's your 45 seconds
the catch is the shoes of course can't reach into your vo2max and alter it
what they do is reduce the effort required for a certain pace, they literally lower required heartrate for an effort by improving economy by a significant amount
but then it's up to the runner to be able to produce and sustain that effort, not all try or can
I've noticed runners with higher cadence get more out of the shoes than those who already run with a long stride-length and slower cadence, this may be due to the trampoline effect where only so much energy can be stored and forwarded and there are limits, more footstrikes = more energy return overall vs long delay in airtime