This has turned into an argument over the definitions of words, at its core. One person is defining "faith" to mean "having a belief that significantly mismatches the evidence", and you're defining "faith" to mean something drastically different. May I suggest either agreeing on a definition for the sake of argument, or splitting it into two separate words? You could call the first definition "mimble", and the second definition "spuzz", for example, and then the discussion might be able to go somewhere.
(I also don't consider your definition of faith to be particularly useful, and I suspect that you're equivocating, but those are tangential to my point here.)
I'm not trying to equivocate at all. I'm just trying to show that 'evidence' works fine in the large -- at the society level. But it doesn't work at the individual level since most individuals don't actually have evidence -- they choose to believe what their culture calls 'evidence'. Call it faith, belief, whatever. But the individual doesn't have evidence.
It's easier to see when we pick something that the whole culture hasn't coalesced on yet. Like string theory and worm holes. Some people believe there's a worm hole at the center of our universe. They have 'evidence' that supports their theory. I can believe or not, but I'll never know one way or another. 100 years from now, if everyone believes the _exact_ same theories we have today, with the _exact_ same raw data, my great-grandkids will say they know the worm hole exists and the 'evidence' is obvious. The only difference between now and then is how widely those theories are dispersed and accepted. Yet today I (and you?) wouldn't say we know worm holes exist. And my great-grandkids won't 'know' any more than you and I do.
In other words, my 'evidence' is what is commonly known and accepted to be true. Yet I personally don't know it to be true -- I trust that it is. Isn't that blind?
(I also don't consider your definition of faith to be particularly useful, and I suspect that you're equivocating, but those are tangential to my point here.)