(Aside: I once peer-reviewed a paper on "pulsatile flow" from another perfusionist. He compared two blood oxygenators for their ability to transmit a pulsatile waveform. He used two different Y axis scales to demonstrate the comparison. A cursory glance made the two waveforms look very similar. A closer inspection showed they were entirely different. He had "stretched" the Y-axis on one of the oxygenators to make it look like it did a better job of transmitting the pulse wave than the it actually did. I caught the "error" and made him change the scale of the graphs so they were identical. The paper was eventually published, but his "conclusion" was also changed. )
So, yeah, the compressed Y axis in the OmniMic graph will make the result "appear" smoother.
Nonetheless, that OmniMic graph you posted earlier was a better response. On it, you were +/- 6 dB from 50 to 100 Hz. On the above graph, you're more like +/- 10 dB from 50 to 100 Hz. And the +/- 10 dB shows on both graphs, whether the Y axis is compressed or not.
Interestingly, the "shapes" of the curves are not that dissimilar:
They both show a pretty smooth response below 50 Hz, the same bump between 50 and 60, a recession in the 60's, and another bump around 70. However, you are currently losing more from there to 100 Hz than you were before. (Of course, the top graph is speakers and subs; the bottom graph is just the subs, right?) Are you sure they are the same 1/24th octave smoothing? A graph that is lower resolution will "look" smoother than a higher resolution graph.