I suspect that the power loss in such an exhaust plume is more due to the elevated air temperature than oxygen depletion. I say this because, in every engine control system I'm familiar with, the control will adjust fuel flow to maintain RPM (or some other selected parameter like turbine temperature) regardless of fluctuation in O2 content.
Your OAT gage in the cockpit may not respond fast enough to give you a true picture of temperature changes.