I fully support the idea of empirical testing. However, keep in mind that there is quite a lot of analytical expressions available for sound transmission loss, even for multi-layer panels. See eg Beranek's classic text books. And, if you don't want to implement these, I recommend commercial...