As of right now, refunds incentivize both the use of gastokens and the clearing of state when possible. The storage savings from breaking gastokens would be miniscule compared to the cost of de-incentivizing the clearing of state.
When a user transfers their entire balance of an ERC-20, they transfer balanceOf(...), that storage slot gets cleared, and they get a refund. However, if there were no refund, they would be incentivized to not clear the storage slot, and only transfer balanceOf(...) - 1, to save gas in case they ever wanted to re-acquire some of this ERC-20. The same logic can be applied to approvals, deposits/withdrawals into DeFi protocols, deployed contracts, and so on.
Given how high gas prices are, and the competition to be “gas efficient” between DeFi projects, most projects would rush to implement patterns that keep storage slots open to save gas. The size of this state growth would likely be significantly larger than the impact gastokens have. In the worst case of no refunds, every storage slot a user touches would be permanently occupied, as clearing it would be inefficient with regards to gas.
Additionally, gastokens are fairly sustainable. They are created with the purpose of deletion. Looking at CHI, the total supply today (1.6mm) is less thant it was on October 8th, 2020 (1.9mm). It peaked around December, at 3mm. The market cap of gastokens will be limited by gas volatility and the number of transactions per block. There is no reason to have excessive amounts of gastoken on-chain, as there is no profit to be had in minting gastokens that probably won’t be used.