Well done on the data set that you've compiled and cleaned Brian - a big part of the task I'm sure!

The main issue with the age grading system in my view is not that someone can score >100% (depends on how we define 100%). It's that the implicit assumption is that if we train the same number of days per week for a decade that our score should (on average) stay constant. I'm sure that's how users of the parkrun age grading system, say, interpret the result. Some runners may alternatively allow for the training regime to get easier over time by the typical amount.

Using percentiles has the same flaw: if our percentile stays the same over time but the population of runners is getting weaker (elite runners training less) or maybe stronger (surviver bias at the older ages) then our performance isn't constant. This can only be resolved by a longitudinal study.

In conclusion, we need to be very clear on what question we're trying to answer before starting to compare methods. Anyway, your analysis is great and your visualisations spot on, many thanks.