Use cases
Stake NFTs should normally not be bought and sold naively. Otherwise the initial owner will lose any rational interest in operating the validator since they don't hold its stake, while retaining the burden to do so. Worse, they have the power to slash it! Hence traders should be aware of what they hold and value it appropriately.
Exit queue
In some cases, if a validator is in the exit queue, selling the stake at a discount will allow the validator operator to access liquidity more quickly.
Lending
If a validator operator wishes to access liquidity while keeping their validator undisturbed, they are able to use the stake as collateral in a protocol such as PWN to obtain a loan against it. In that case, borrower will be assured that if the validator defaults, they will be able to seize the collateral and extract what's left of the stake to cover their loss. Assuming adequate parameters, the validator will be incentivized to take care of their loan and avoid seizure.
Staking Derivatives
A special case of lending arises when ETH itself is being borrowed. Price fluctuations are no longer possible, so the only source of profit or loss on the collateral is the performance of the validator. Assuming there is still a comfortable buffer between the total stake and the lent amount, we have created a situation where the lender earns staking-derived fixed income with near-zero staking risk, and the borrower is running a validator with only the buffer as their own capital. We have created a trustless staking derivative! Additional work can allow the validator itself to be bootstrapped from a loan, hence allowing the operator to enter a leveraged staking situation without ever holding 32 ETH. Furthermore, such loans could be aggregated, syndicated and fractionalized, hence creating a form of fungible staking derivative.
Interoperability
Suppose there is an LST protocol with an oracle that allows operators to spin up validators with 4 ETH of their own paired with 28 ETH from other users. Suppose this protocol is built on top of ST. It could be made possible for the operator to transition to solo staking without having to exit the beacon chain simply by repaying the 28 ETH and receiving the unencumbered token in exchange.
L2 deployment
The architecture of Stake, Tokenized makes it so that deploying the core contracts on L2 is possible with some modification. Of course, some code must be put into the L1 state tree to move the ETH, but ownership information, access control logic, and integrations can stay on L2. Even an L2 with relatively weak composability with L1 is enough since contact between the consensus layer and L1 execution layer is also asynchronous.