TLDR; DeFi protocols would be some of the least efficient banks in the US.

I’m a big believer in DeFi, but people have to understand it’s still very primitive and inefficient. To understand what I mean, let’s zoom in to a key metric for lenders - the net interest margin (NIM).
NIM is a measure of how financially efficient a bank or other lender is, and is calculated with the basic formula of (Interest Income - Interest Expense)/Average Earning Assets. We won’t clog the timeline with calculations here, but some quick estimates on NIM for major protocols:
Aave v3 (Etherum): ~0.4%
Compound v3 USDC (Ethereum): ~0.9%
Maker/Sky: 1.9%
This means Aave v3, Compound v3 (USDC), and Maker/Sky would be pretty low on the bank rankings by NIM.
There are about 4,400 banks in the US. Were they banks, Maker/Sky would be 4,274th, Compound v3 (USDC) would be 4,344th, and Aave v3 would be 4,346th.
For another comparison, the average credit union in the US has NIMs of around 3.3% right now. If the average credit union was a bank, they’d be around 1,500 places higher on the bank NIM list than Maker/Sky, the protocol with the best NIM that we’ve looked at.
And remember that credit unions are nonprofit organizations that are NOT trying to maximize NIM for the benefit of shareholders!
Hopefully you can see why investors in lending companies and banks are not especially impressed by even major DeFi protocols.
There is a silver lining here. It means there’s a lot of room for these protocols to grow. That, of course, requires having a better understanding of the financial performance of protocols, which is discussed more rarely than one would assume, given a key advantage of blockchain is that a project can measure its performance in real time.
It should also give pause to the current operators of these protocols. While no one expects absurdly high efficiency, DeFi projects have minimal regulation to constrain things like the mix of loans they originate or govern liquidity requirements.
That even some of the most well-known, battle-tested, and heavily utilized lending protocols are operating like a Model T instead of a Tesla indicates there needs to be better focus on core competencies and growing the product mix to serve their existing user bases.
It also suggests that current management at all three need to bring in some fresh blood with new ideas or consider stepping aside.