Indeed, there might be a qualitative change, but no quantitative one for node runners. The change under discussion at the moment will probably not lead to a big change in data embeddings in the chain though. Stamps, brc-20, and inscriptions will porbably not move their infrastructure to a potentially more expensive OP_RETURN setup. Bare multisig, and emebedding data in the witness will probably remain more efficient for them. The positive hope of this change is that future protocols that need to anchor data in the chain every so often, will do so with in a prunable, and non-UTXO set polluting way. They will embed the data anyway, but changing the OP_RETURN limits might reduce harm.