It is also currently possible to pay transaction fees off-chain on virtually any blockchain using fiat via SWIFT transfers or Western Union payments (as tongue-in-cheek examples) to prospective blocksigners in exchange for them to include your 0-fee transactions in their blocks.

But this is obviously not practical/efficient for large-scale implementation, and especially not as the default/intended user behaviour on a network; the model you propose is only marginally better in this sense (and is worse than the aforedescribed off-chain method in terms of blockspace efficiency)

Furthermore, I think you missed the central part of our proposal, and what makes the "No Coin" principle we described so important in the first place:

We want to optimize the sidechain for cross-chain swaps with on-chain/Lightning BTC. This is achieved mainly through the "Anchoring" consensus rule I described in OP, which is arguably the most important difference that we're proposing over the current version of Liquid.

In other words, the "No coin" principle is necessary mainly because we want no part of the sidechain protocol to be technically dependent on the existence of pegged Bitcoin, so that the sidechain can outlive the usefulness of pegged Bitcoin; but the Anchoring rule is the primary innovation that makes this possible (creating cross-chain consistency between the Bitcoin timechain and the Sequentia sidechain so that swaps and other HTLC don't require burdensome timelocks to plan for edge cases like Bitcoin chain reorgs).