Currently, S1199 reports an issue on any block in a case group. This approach is too naive, and result in issues with actions that do not make sense:
In fact, this rule should not report an issue when the body of the case is a single block. If there is something to refactor, RSPEC-1151 will report it.
This includes the new way to write switch in Java 14:
Still, when the body contains multiples statements and one of them is a block, the issue makes sense: