TOPIC: UMIK-1, REW & Car Audio

Good Morning All,
I'm having a bit of a tough time with using REW to help me EQ my car audio system. Whether I measure one side at a time or one speaker at a time, I'm getting some wacky EQ filter suggestions like boosting 9db at 2100hz with a Q of 5 and the reducing 9db at 2250hz at the same Q. I suspect it may be the manner in which I'm performing the pink noise test. But let me detail what I've been using:

UMIK-1 microphone and the non 90 degree calibration file (mic positioned at headrest facing forward)
Macbook Pro with REW V5.1
Audiofrog House Curve
Pink noise at 0db with my iphone connected via CarPlay into my car audio system (my most commonly used method of music source)
Volume set at my normal listening level

First off, the test seems to yield a proper frequency response but it's far different in db than the house curve. So I apply an offset to get them closer and then things go wonky from there. I believe my problem lies with this offset or with the fact that I'm not using REW to generate the test tone. Should I just be using the headphone jack out of my MacBook Pro to the USB input of the car stereo and REW's test tones instead? Or does that introduce the laptops soundcard?