Most microphones that we use and the ones for room measurements, are omnipole. Most manufacturers give eq curves for on-axis and 90¤, but difference is minimal. So is the difference in long IR gating measurements with the mic pointing at ceiling or speaker.
Eq'ing sharp dips is not possible or wise. They come from destructive interference/reflections, and the higher signal is, the higher is also reflected spl, resulting in null again, but if we move the speaker of mic, the dip's frequency changes moves and we have a huge bump in response.
The 720Hz dip is most likely a first reflection too - sofa or laptop or even side wall. The source is easy to find and fix by doing several measurements by moving mic, furniture, speaker etc. one per time. It also might be L/R comb filtering because both speakers were playing in the measurement.
The general (or better, majestetical) rule #1 is - thou shall not try to eq sharp dips in response!