I completely get what you're saying, I think you're missing what I'm saying. The problem is that this only works if you're able to commit 100% to either holding through delivery or only trading during roll days. If someone is in that situation, then perfect product, for sure. But if there is even the chance that you'll have to exit a position any way other than a roll day or through delivery you'll destroy years worth of bps when you have to cross the spread, no? That's makes the risk adjusted return pretty much negative for anyone except the investor who knows with absolute certainty that they will carry through expiration or trade on a roll day. You're effectively doing a lockup, and the market generally demand a lot more than 19 bps for a lockup.
Ah, there are ways to accelerate expiries. Exchange of Future for Physical was the historical transaction for this and it remains available but severely limited. We use to offer them but the SEC didn't like it so we came up with a new transaction. Securities Tranfer and Return Spreads or STARS. Each day we have a SSF that expires with T+1 settlement. When combined with a longer dated contract it allows for the acceleration of the longer dated to today. Since the STARS is just a riskless transaction (the spread provides the hedge) their are a number of financing desks who would be happy to accommodate by offering at a reasonable interest rate. For those familiar with the IBKR EFP offering you will note that they no longer offer EFPs. The STARS transaction with the T+1 settlement of the front leg is a superior trade and they have abandoned their EFP offering altogether.
The issue with STARS is that depending on the symbol they either aren't actively quoted, are at a wider spread than the outright SSF, or aren't offered at all. Have you considered dropping the market for outright SSFs and forcing everything to be done through spreads/EFPs? I'd expect those markets could be much more competitive as they don't require any HFT infrastructure to make markets, but right now all the retail order flow goes to the outright SSFs so there's no incentive to do so. It seems to me the main advantage to SSFs for a retail trader is earning the cost of carry of when shorting easy to borrow names. On many names this can mean an extra 250bps. Hard to borrow names are great in principle, but there's never a market there. For financing a long position, the benefit is much smaller as you pointed out, and you can do even better selling a box spread.