I believe the warning you are referring to in the viewer says - "The underlay/overlay pair of datasets have oblique angle difference of X.XXXXX degrees. This may cause them to appear out of alignment in the viewer. If you are performing spatial transformations on an Oblique dset, or viewing/combining it with volumes of different obliquity you should consider running 3dWarp -deoblique on this and other oblique datasets in the same session."
The warning is to let you know that your datasets differ in this way and that (along with other reasons) may be why the overlay and underlay datasets appear out of alignment. You can mute these warnings if you wish.
align_epi_anat.py can change the orientation of your EPI to match your high-res anat if you add the flag -epi2anat to your command. Also, afni_proc.py will correctly perform slice timing correction on the original data (oblique data if that's what you have) before aligning the data to an anatomic dataset. You can see this in the output of afni_proc.py by running:
3dinfo -prefix -orient *.HEAD
Which if you have volreg_align_e2a set, shows that the EPI and ANAT start off in different orientations but end in the same orientation. Also you have the option to have the anatomical data aligned to the EPI.
I also wouldn't assume that all fMRI practitioners come from the same background.