You're adding a bunch of risk for literally no upside.
The only thing you gain is to to trade that on the blockchain, which is really a response to the oracle problem.
So you've greatly increased your risk with no upside to slightly increase your utility. That doesn't seem like a good deal.
Stable coins can be exchanged against other crypto currencies on chain. So if you are willing to trade crypto currencies, stable coins are more practical than USD.
There are also tax implications: Depending on where you live, crypto-to-crypto profits are not taxable. You will be taxed if you sell crypto currencies for USD, but not if you sell crypto currencies for a stable coin.
Does that also mean if a stable coin fails you can't write off losses?
At least in the US, this is not true.
e: Not sure why I'm downvoted - crypto to crypto conversions are taxable in the US.
I think the "investing" in tether is really only the best match to "Let me see, where can I park my money in the least volatile way on this crypto exchange".
EDIT: Interestingly, if tether breaks, then USDC is the last man standing and will have to work hard to maintain the peg. Under such circumstances it's quite possible that BTC is the next best "stable" coin.
Either way, most stable coins are built on some form of trust, and I'm much more inclined to trust Coinbase than most others. (Disclaimer: I have no dog in this hunt, I don't own any crypto whatsoever.)
Unlike Tether and USDC, I can actually verify that Dai is overcollateralized.
Well, yes? Isn't that enough to explain it? Normal US electronic dollars can't interface with Blockchain smartcontracts. So if you wanted to invest in e.g. a vault or liquidity pool, you would need to turn it into a stablecoin first.
You mine a Bitcoin that 'should' be worth 5k USD, but is instead selling 60k USDT. So instead of selling that Bitcoin for 5k, you instead sell ~8% of your Bitcoin for 5k USDT, pay your expenses, and keep the other 92% of the Bitcoin you mined.