you're right, we tend to place too much value on what objects 'look like'...
so what i'm doing is to try to make a case based on something OTHER than 'looks like's,
while attempting to get past some of the shortcomings of the 'static' nature of photos vs.real life as you mentioned above.
the process is actually inspired by the land overlay mask you posted elswhere on the board.
it concerns itself with the mathematical relationships between facial features. instead of making a mask of those proportions,
i used the 2nd photo as the mask...
after carefully placing the eyes of the 2nd pic in EXACTLY the spot the eyes are in the 1st photo, then adjusting the opacity.
its essentially a poorman's land overlay mask.
the problem is i haven't presented it near as clearly as you do.
the screenshot moved, images didn't crossfade smoothly,
and i used the word 'morph' instead of variable opacity crossfade.
this new video should help:
www.youtube.com/watch?v=fRtzCEmfWOsi just realized it just looks like i'm flipping thru photos.
so i've worked in a variable opacity crossfade.
on the first pass, keep looking at the eyes, you'll see that they are ALWAYS in the same spot thru and thru the video.
on the 2nd pass, concentrate on the nostrils, bend of the nose, distance to lips, shape of mouth.
as a final touch, i then ADDED the land mask to underscore the sameness of the images.
once again, the land mask will always appear in the same spot
in the video.
(ok, you will see that i tried to make a mask myself
but it came off looking like poorly-applied paiute war paint)
rather than dealing strictly in 'looks likes'
or 'flipping thru random photos' each of those photos had some connection to our subject.
actually my 'looks like' radar told me some of these were not him, then to my astonishment they matched this well!
and the mcCarty portrait, albeit the least proven image,
is being hailed as brewer instead.
(sorry steve-- love your work and you got a lot of things right
but this just isnt him)
he even posted next to a pic of brewer
and after having read land,
for the first time ever i could instinctively see past 'looks likes'
and immediately read the 'imperfect cross' of facial asymmetry as a non-match,
then reversed one image in my head for flippage and i could tell it still wasnt a match,
tho they did seem to 'look like' each other.
jokingly i said the asymetry matches billy instead,
then to my astonishment it really did!
i DIDN'T use the globe photo 'cause it didn't match quite as well.
in conclusion, i guess this stuff really doesn't prove anything, there isn't some high council
which determines the merits of photo id's,
and its challenging to convince anyone
b/c it just 'looks/doesnt look' like them in perceptions.
but... there maybe someone else who matches like this,
but try to find me 2 people who will, from the same era like this.
i can't resist dabbling in photos,
its something we can effect immediately.
but you're right, there's evidence of other forms
which may be deemed of more value,
just not something i can immediately effect.
www.youtube.com/watch?v=fRtzCEmfWOs