Well, one of the "advantages" of formal testing first for the teacher is that often the classroom teacher doesn't feel like they have to do anything while waiting for the consent-schedule-test-report-meet-plan-implement process to wind its way. Plus, it's now someone else's responsibility to figure out the student. And then, let's not forget that it keeps people like me employed!
...To be fair, I spend more time than I would prefer talking people out of testing, and it's not limited to teachers and other school personnel. Plenty of parents and outside careproviders (physicians, therapists) reach for formal testing first, too, without having a clear idea of the actionable information that they want to obtain. Don't get me wrong, I think thoughtful, thorough evaluations can be extremely valuable, but they're best executed when you know what your question is, and in the context of rich naturalistic sources of data, such as the aforementioned informal assessments.
To the OP: I believe that testing is warranted when there is a question/problem to be addressed, or when it is needed for access to resources. There have been many thoughtful responses upthread; I agree with many of them in that mutual discussion and understanding of you and your partner's expectations regarding testing, its functions, and its possible outcomes may go a long way toward identifying areas of consensus. Data informs decision-making, but it does not mandate any particular choices.
