Comparing ratings between generations is completely useless. Ratings are based on the players in the pool, the amount of players in it, inflation, the way ratings are calculated, and a host of other things besides playing strength. Ratings can only differentiate between players within the same pool at exactly the same time.
But even then ratings are not a good method of judging who is the best because they don't objectively evaluate solely playing strength. Ratings are in big part based on a players consistent performance. Someone could be 2950 strength in normal play, but if he was known to have periodic terrible migraines a couple of times a year that reduced his playing strength to 2550, his rating would settle somewhere far lower than his actual strength. Similarly if someone is a really strong player, but nerves, stress, or tiredness, stop them from playing well in all of their tournaments, ratings will never accurately show their true strength.
A similar analogy would be baseball regular season records. It's often not the best team that has the best record, but a good team with a very low number of injuries that takes the top spot. That's why baseball doesn't use regular season records to determine who wins the championship. They take the teams that have been shown to be pretty good by being filtered out through the regular season process, then have them play each other to determine who's actually the best.
Similarly with chess, tournaments and rating can be used to reasonably determine who the really good players are, but you need matches to determine whos the best, since purely statistical means like ratings or records aren't good enough. Thats why Carlsen just having the top rating wasn't a good enough reason to say he was best in the world, it was when he beat Anand the world champion (and number 3 rated at the time) that he showed he really was the best.
Agree