Absolutely, that is true, yes. Market making cannot help enforce the peg. It can only add liquidity around the price level where buyers and sellers are willing to meet. The only thing that can help strengthen the peg is convertibility and the ability to arbitrage in free markets around that. And that depends on the price feed. This is why I criticise the view that the pegging problem will be resolved one day once we have enough liquidity in the market. I can expand further if you like.
Doesn't this mean there is little point to having an internal bitAsset market? Couldn't you just replace the whole thing with a system which created and destroyed bitassets directly at the feed price, as long as each side had sufficient collateral?
You got me thinking laterally here. Though I'd like to ask what you specifically mean by an internal bitAsset market?
Let's consider currencies like bitUSD. The bitUSD needs to be an on-chain token to preserve its superior transactional features. And derivative markets of the bitUSD such as bills/bonds, credit markets, options and futures etc, would all be best placed internally to provide the superior exchange features that Bitshares has to offer, as well as to maximise coordination between these markets.
Specifically for the creation and destruction of bitUSD, I believe such exchange should occur at the feed price, and there should be an official window where this occurs (in my white-paper, that is the Currency Creation Market). But if we are willing to consider a more open architecture to the creation of the bitUSD itself, yes I suppose you are right that the official window where bitUSD is exchanged for collateral could be external to the system, such as a UIA, with bitUSD as a UIA token. I expect this adds additional risk however, and may not be an optimal path. My preference is still for this window to be internal.
As we are discussing elsewhere, its also possible that the collateral used is not BTS. It could for example be BTC, currently the most recognised and accepted form of digital collateral. This could be converted to an on-chain collateral token (e.g. a substitute bitBTC) useful for many other bitAssets and applications, but again with extra risk that needs to be minimised.
It's further possible to consider alternatives for the structure of the short side. While it makes sense for short speculators to be on the other side of bill, bond, and derivative markets, where there is greater symmetry between longs and shorts and markets have a floating price, at the official window as I've described it might make more sense for the short side to consist of a pool of investors sharing in the benefits of currency issuance, such as a spread on all window transactions.
From a purely commercial perspective, these are all possibilities that can be considered and the various pros and cons weighed against each-other.
So yes, once we reject the idea of an endogenous price feed, and accept that the price feed is critical to underlie convertibility and valuation throughout the system, then the thinking naturally leads in some of these directions. I think we can consider a product suite better suited to current market demand as a result.