You're making an assumption here that I would regard as dangerous. My main problem with this is... okay, so you fly a 5-degree offset as indicated on your ADF. How do you KNOW FOR SURE that your ADF needle is properly calibrated? If you actively insist on going at least 5 degrees OFF the published track, you have just eroded ALLLLLLL of your regulatory tracking tolerances. Is your ADF needle really sensitive enough for that? I would counsel some caution on that, if you're flying raw data.
I fear I'm not making my point well, for lack of a diagram. If, instead of adjusting the o/b track, I fly the published outbound and accept that the turn will roll out past the final approach track, then at the moment of roll out I will also be 5 degrees off the final approach track. Isn't that equally "ALLLLLLL of my regulatory tracking tolerances"? Why is that in some way better than being off track on the o/b leg?
One advantage that I have in making the correction on the o/b is that I have course guidance and, within the accuracy of the instrument, I know where I am with respect to the published track. I can make sure that I'm no more than 5 degrees off. By contrast, if I fly the published o/b and accept the drift in the turn, the first I'll know about how far off the FAT I am is when I see the needle indicate it at the completion of the turn. For all I know, it may be 10 degrees off.