you could also try inverting the phase on one of the two mic channels to see if that improves matters.
then I would suggest using track delay (in the inspector) to offset one of the tracks playback by a few milliseconds to see if this gives you better phase correlation. (perhaps loop a section and adjust these settings during playback till you find the sweet spot)
once you’re happy, you can then render the modified track and check the transient alignments visually too if you want. (but personally ears>eyes)
hope that helps.