Yes those confidence levels should be fine. Only 3dB between L & R main speakers means your mic was fractionally off centre so look at the impulse response of both in one viewer window, zoom in and take the mid point timing as the reading. (It will probably be no more than 1 or 2 fractions of a ms difference between the two main peaks & only really visible at a high zoom level).
The best way to set mains to subs is first to align the faster sub (usually nearest to your listening/measuring position) to the slower sub, then align the pair of main speakers (average reading as above), also with the slowest sub.
Finally, take another measurement afterwards to verify & you should see both subs initial peak at the same point & the main speakers main peak now also at the same time alignment. I always find it best to open two viewer windows, one for the pair of mains, one for the pair of subs. If you try and see all 4 measurements in one viewer, the scale usually makes the sub measurements look almost flat and impossible to interpret accurately.
The best way to set mains to subs is first to align the faster sub (usually nearest to your listening/measuring position) to the slower sub, then align the pair of main speakers (average reading as above), also with the slowest sub.
Finally, take another measurement afterwards to verify & you should see both subs initial peak at the same point & the main speakers main peak now also at the same time alignment. I always find it best to open two viewer windows, one for the pair of mains, one for the pair of subs. If you try and see all 4 measurements in one viewer, the scale usually makes the sub measurements look almost flat and impossible to interpret accurately.