Frequency response curves

Here from one of my 1988 log books is a traditional pen-chart trace of a speaker under development. If I couldn't recall which model this prototype would become in production, I could not deduce its name from this curve. Frequency response curves require such an intimate knowledge of the measurement set, the microphone, minor squiggles in the trace (reflections from the measuring environment?), where the mic was placed relative to the drivers etc. that, on their own and without any of this detailed knowledge, meaningful interpretation is completely impossible. They are as good as useless. Pretty though. Can you guess where the crossover frequency is?