Suppose something like a scheme for writing blobs to the blockchain using naked OP_MULTISIG calls takes off big time and the UTXO set starts to grow a lot, how can Bitcoin nodes defend against that?
I imagine it is possible, for example, to flag some types of UTXOs as much less likely to be spent, moving them down to the depths of the disk, while leaving other UTXOs at the top so they are kept in memory.
Another simple heuristic that comes to mind is just the age of the UTXO -- older UTXOs are (probably?) less likely to be spent than newer UTXOs, and so on.
Are there other of these heuristics that can be applied? Are any being applied today by Bitcoin Core or some other Bitcoin node implementation?
