Notice the above proposal does not increase the supply of BTC, it treats all BTC the same regardless of it's history/owners, and it does maintain a very good incentive to mining.
Economically speaking, your proposal is really no different to raising the supply and having an endless emission of bitcoin.
Bitcoin will have endless emission regadless, we do not know what the value of transaction fees is going to be, it could be greater than the current block rewards and thus more hashrate could be running at a profit then.
In your proposal you take a fee of (for example) 0.5% from every coin every year. Each year, everyone loses 0.5% of their coins and gets 0.5% poorer
.
If the rewards from fees are not large enough to keep enough hashrate securing the network, then everyone who owns only Bitcoin will get 100% poorer.
Alternatively, by raising the supply and minting an additional 0.5% of the cap each year, then everyone owns the same amount of coins, but a smaller proportion of the overall supply. And so each year everyone gets 0.5% poorer.
The end outcome is the same regardless of which method you choose.
The outcome is far from being the same, one main point of strength of Bitcoin is its finite supply, raising the supply past 21M even if it was 1 coin a year will make the supply infinite just like all the fiat currencies, the economical difference between that and circulating a tiny portion of BTC from hodlers to miners is entirely different.
By rasing the supply Bitcoin is guaranteed to lose value, while with extracting fees you just distribute a small portion of that values between different addresses.
Ya I personally wouldn't want to pay fees just to store my coins, but it is also naive to expect that my coins will be secured enough when nobody has any motive to point hashrate at the network, when a few dozen double spending attacks are done cheaply without an issue, everyone will be glad to pay a small fee.
The alternative to that would be random people mining at a loss just to maintain the network, that could work but difficult to guarantee.
Now all of that is probably not going to be needed given that in 60-80 years time you would expect the transactions will be a lot higher in number as well as value, but you can never be too sure.