I very much agree with this. We use a lot of formula names but there are a lot of downsides. Ignoring the difficulty with creating entities w/ formula names (talked about in many other threads), another huge downside to formula names is that making entity names distinct naturally makes the name longer.
For example:
Vendor-A works on a lot of my projects. The way my schema is setup, every time Vendor-A is assigned to a project, a sub-entity is created. “Project-A_Vendor-A”, “Project-B_Vendor-A”, etc. Having distinct sub-entities per project is very important to collect information about the vendor specific to that project. And on the original Vendor-A page, I can aggregate information across all their project pages.
The problem, like @arthurpires noted, is that if I link their project page anywhere, the relation field says “Project-X_Vendor-A”. It ends up being a very long title. And when the View I’m looking at is project filtered, the “Project-X” portion of their name is no longer relevant, it’s just taking up valuable space.
But alternatively, if you take away that portion of their name in the formula, it makes it impossible to find them in search or a non-filtered relation list.