Now, the notion of "match case" in find&replace is rather ambiguous in itself, but why don't we simply have in this costly translation software what MS Word has offered since stone age, namely performing replacement on a 1:1 basis, i.e.:
If there are phrases like "Disabled Persons" and "disabled persons" in a project (aka my current project) and I want to replace it with "disabled people" PRESERVING THE ORIGINAL CASE (which is around 95% of real-world user scenarios, or isn't it, Studio engineers?), the software should do exactly this as default (or upon triggering the Match Case setting) and not overly complicate things.
SDL has recently gotten "inspired" (aka copycated it) by one great Word feature which is smart cutting and pasting, so why not go one step further and introduce this feature in the next SR.