If that is true, then the units might not be rare but blockchain space might indeed be very scarce (and then must be somewhat expensive). It is a useful perspective.
At 8 billion people and 1024 account balances each, and say 128 bytes per account storage, that is 1 Petabyte. That is the best the mini blockchain could do. One-time ring signatures would make it entirely intractable (100s or more of Petabytes) unless you increase the cost until the 8 billion are restricted.
Let's assume a 10 year goal of 1 billion x 8 account balances each, so 1 Terabyte which is one hard drive. And hard drive space doubles every 18 months, so in 17 years we can store 1 Petabyte on one hard drive. So the mini blockchain can scale.
So there is no scarcity, except
for badchosen design (priorities).
P.S. we do have to discourage dust balances though.
No one has demonstrated how to do privacy with a mini-blockchain (mini-blockchain is only minimally demonstrated, but at least that's something). You snipped the important part of my post, which is that scarcity may exist if other priorities conflict.