This would be useful for Search 6 in particular, to know which parts of the current code are actually tested, and which parts are not and are thus potentially useless.
It wouldn't give definitive answers of course, more investigation would be needed, but it would at least give us some leads.