Okay, I'm a chemistry major and it's been awhile since my last bio class, I'm certainly no genetics expert, but let me take a crack at it.
I think we may be looking at this all wrong, we presume that there are only two traits; the photosensitive trait and a photosensitive inhibiting gene that suppresses aforementioned trait. The photosensitive trait being dominant and the photosensitive inhibiter gene being recessive (or whatever the fk it is, lol), but I believe the reason we are seeing such unexpected results in F1 is because we are looking so deeply into the 'auto' trait, when I believe the answer lies in the dominant trait... Or should I say 'traits'?
I believe that it is possible that these are several photosensitive traits; early flowering, late flowering, etc. and that they are independent of one another. I believe that some of these traits are dominant while others are 'partially dominant' and that is why sometimes we see the ratio that we expect in the F1 and sometimes we see a ratio that we do not expect and get mixed results. The same could go for the 'auto' trait, is it not possible for there to be multiple genes that display the 'auto flowering' trait?
In the end, I have no idea wtf I am talking about and am not qualified to speak on matters of genetics as I have never breed a damn thing in my life, lol.
It seems logical, from an evolutionary standpoint, to have this 'auto' trait present as it makes the plant adaptive, although it may not be necessary at times, at other times it may help the species survive and reestablish photosensitive varieties in the future, after some strange global phenomena. In my mind it just increases the species survivability as a whole. I know that sometimes we have longer or shorter seasons, based of the earth's orbital pattern around the sun, could the 'auto' trait be a built in protection from extremes in seasonal changes?
... Please be a gentleman when smashing my makeshift hypothesis to pieces, lol.