Interesting. This topic has been discussed previously - that very interesting "raw" data.
My "rate" of requirement in my personal experience is certainly way higher than 1/600 flights - about 1/6 at one stage! There will no doubt be a host of factors that make certain flights "higher" risk of medical emergency (perceived need to travel between destination combinations, perceived "risk" depending on journey difficulty, flight length etc), and aircraft size would also potentially increase the odds to some degree too (although also probably increases the odds of one of the pax being medically trained!). The journey scheduled duration would also impact diversion rate ie. 1hr out of LAX on a LAX-MEL when an emergency occurs is much more likely to "divert" (back to LAX) than 1hr out of MEL on MEL-SYD!