If Splinterlands solved ratings inflation and accounts were in leagues appropriate for their card levels, then yes, the incremental increase to get to Gold league cards is on par with the incremental increase to get to Max level cards.
However, because ratings inflation allows bronze level decks to get to Diamond league, that definitely distorts the cost/benefit analysis and what is required to get Diamond league chests.
That's why I keep emphasizing why ratings inflation is such a major contributor to the underlying problem.
I think that you and Bulldog make decent points about the way rating works currently is broken, but I would argue that we don't have a ratings inflation problem. There is no market cap on rating.
What we have is an issue with how much rating affects rshares, which makes rating more important than winrate, cards used (even with the GF and Alpha/Beta/Promo bonuses), and how nothing replaced the Airdrop mechanic where holding assets rewarded you with more assets.
To me, the issue is less about rating and more about building out a more complete and rshare algorithm that rewards and does not punish on many factors that would drive bots to behave more like typical players in holding assets on the account and using combined cards.
I don't think we can solve this with just messing with rating. We need something better like prop 1254.