There are very many people getting OOPs something went wrong on an almost constant basis, it appears to be a new built in additional feature of the new and improved search process.
I'm just an ordinary person, but here's my offer of a possible reason for this error message.
Rather than write a long list of descriptions and index numbers for all possible errors that could occur in a query, which, let's face it, would only be of any use if you wanted to fix it, it's a heck of a lot easier (and cheaper) to just write one message for everything. That way the implementer can argue "Well I can't fix it if you can't tell me what it is", and get back to playing SnarkQuest®
Background: Many years ago, I wrote an extensive error trapping and recording procedure. When the system went live the printer starting spewing out reams of paper at an alarming rate until somebody hit the kill switch. I had included among the errors a number and description for every instance where a command was invoked after the associated action had already been taken. Investigation revealed that the SW engineer had decided (i.e. it wasn't in the spec.) to make every, yes every! command occur 5 times, "just to make sure it happened". Result: Lots of waste paper. Perhaps word of this event dispersed through the mythic mindspace of SW engineering to create a general approach to ensure that such an embarrassing situation could never happen again, by just having a single non-specific error message.
" DON'T PANIC ! "