I think I still haven't stopped laughing at the idea of Cavendish being under-rated. Probably THE most over-rated cyclist of all-time. Take him out of the modern sprint train era and he might have been a slightly better version of Jean Paul Van Poppel.(This can also be applied to any sprinter of the modern era of course).
If he had raced in the Van Poppel era(85-94), would Cav have a World Title? Not a chance. Would he have won Milan-San Remo? Nope. How do I know this? Well, because neither of those races ever finished in a bunch sprint during Van Poppels career.
Would he have won 20 stages in 4 years at Le Tour? Not a chance. If Van Poppel had managed to win every single sprint finish stage at the Tour between 1987-92, he still wouldn't have 20 vivtories. If he had won every single sprint at Le Tour during his career, he still wouldn't be at 30. So the difference is far less about talent and more about opportunities and of course it is not the sprinters who create the opportunities, but their teams.
I don't understand this fawning over sprinters simply because they get way more opportunities to win than any other type of rider. Is Cavendish a much better cyclist than say a TT rider like Tony Martin? Both specific cycling skills, but one might give a rider between 10-12 opportunities per season max whilst the other......well. A max of 2-3 opportunities per GT whilst having to compete against the top guys in the race versus innumerable chances against a limited set of opponents. It is not the same thing at all. Yet I don't think a single person would consider Tony Martin for the list, so he is disadvantaged purely on the basis of where his particular talent lies and the fact he doesn't get the same opportunities.
When people talk about the greatest sprinter of all-time, they really mean the last 30 years because the sport was so different before then. When we compare Merckx v Pogacar, we are at least basing it on the same races even if you think the level was lower in the 60/70s. With sprinting, it's all about the number of opportunities which are demonstrably very different.