There is another way to look at it. Track level has a lot of control as to what is sent further on. Imagine 8 tracks. One or two of them will have the sounds that are going to be the major par part of the bottom end of the final mix. Say kick and bass. The other tracks might have some extraneous low end noises on them, rumble, deep subsonic stuff and even a slight build up of bottom end of the sound that may be on those tracks.
The total sound now reaching the final stereo buss will have the two important tracks comprising your bottom end sound plus a whole lot of extra low end stuff that could be easily seen as unnecessary.
If you are trying to remove it at the stereo buss you sort of can except that in order to clear it up you might have to interfere with the other two tracks. That low end EQ curve may now suit the cleaning up duties but not favour the bottom end of the music such as the kick and bass sounds.
Filtering all the unwanted stuff at track level ensures that the only important bottom end now that reaches the stereo buss, the two tracks that are much more important low end wise. The sound there will be cleaner and punchier without any extraneous stuff going on at all.
In mastering you do actually apply some high and low pass filtering on the mix at times. If you only have the important stuff in the low end present in your stereo mix now when and if you do use a HPF you can set it quite differently now to do just one important job and that is fine tuning the low end of your mix rather than having to set it for clearing up all that extra subsonic stuff. The sort of EQ curves you are playing with now are focused on just the kick/bass stuff going on. eg leaving the deeper frequencies in so that bass/kick can still sound fat at 41 Hz. (bottom E)