Other than that, I think there are other issues. For example, compared to existing alternatives such as using a multi-sig to secure the loan (if it is a collateralized loan), your solution would add another insecure layer. If it is an uncollateralized loan, then the lender can easily solve this by limiting the criteria for borrowers. While your solution can help them reach other lenders, there is not enough incentive to take that risk compared to what we have right now. At least that's how I see it when I lend to somebody here.
There are a few risks with current lending contracts on BitcoinTalk. First, even though both parties use a multisig wallet to secure the collateral if the liquidation price would reach, the borrower may decide not to share the signature just for the sake of argument. In addition, it's not giving the rights to lender to scam the borrower if he decides to sell the collateral immediately after receiving it to make a 30% profit.
Your idea is similar to Kanpeki then. Last time I checked, not that many users are using the to borrow/lend, probably because most of the users are just looking to earn more profit by depositing funds to some lending pool instead of manually finding contracts and funding them. Try to check them out and see if you can improve on their model, who knows maybe you can find customers for alts lending.
Kanpeki Finance is still more B2C, but I'm talking about C2C (How currently Zenland Escrow Platform operates). With every new lending agreement, 2 users deploy a new contract into the blockchain.
But I agree that there's not enough demand for services like this. As you said, it's easier for users to deposit their money to some DeFi business and get their interest rates rather than finding people directly on forums and creating contracts with them.