The answer to this question is one of the "hidden superpowers" of Ethereum's rollup-centric roadmap. Institutions have a choice - they can have their own chain, OR they can simply deploy their product on a pre-existing general purpose L2. Robinhood, for example, have already launched tokenized stocks on Arbitrum One - they have also expressed intention to spin up their own Orbit chain. L2s give institutions the ability to "test the waters" by deploying on a general purpose L2 first, at which point they can assess the demand for their product, and see if it makes sense to deploy their own chain. They may also require a lot more customization to allow their product to better suite the needs of their customers - that's another reason why app-specific L2s may be required for some institutions/builders.