Either the newly found suspect (who supposedly confessed) is the spitting image of Mr. Grant or the original Crime lab and
a whole bunch of eyewitnesses were horribly mistaken. Maybe they are just liars.
How can this man ever be compensated for spending the last seven years of his life in a cage?

identifying people not of a different race. So it would be interesting to know how many black people were on the jury.

The race of the witnesses. The area of town this was in was not a black area. The bar that it happened at was a gay bar. So it is likely most of the witnesses were not black.

I don't know how my mind switched "witness" to "juror."

It befuddles me.

IMO they pressure witnesses so much to go after someone after they get a "hit" that they lose sight of justice. It's very possible this guy had a bad reputation and stuff like that which is another motivating factor for cops to get guys off the street by manipulating people. Witnesses are not reliable at all. And they themselves can believe bad information and the police can believe them to and even the police may not know that they're even doing it. I'm not necessarily saying it's an intentional process.

They make mistakes all the time.

What a surprise.

Sounds like the lab came back with nothing identifiable. If the lab had reported it was Grant that would have been one thing. DNA tests can be wrong, and should be run twice. On the other hand, test results sometimes come back without useful results.

In this case, it was witnesses. Witnesses make for bad witnesses. They misremember because they didn't see, but since they think they do they construct the memory. It's worse when you're not sure and somebody suggests that you become sure. When you have to struggle to remember, you're likely getting it very wrong, esp. when others assist you.

Witnesses re-remember--every time you remember something, you run the risk of overwriting the original memory. You don't know you're doing it, but you do. You relive the experience and each time it's a slightly different one. If you take a drug that blocks memory formation for 24 hours and during that time are asked to recall something that happened weeks or years before, no problem--there it is, the memory. If you follow up two days later and ask the person to recall the same memory--it should be fresh, right, you just recalled it two days before--there's a very good chance it'll be missing or blurry. Because when your brain went to rewrite the memory, the signal was corrupted. Instead of rewriting it, you wrote nothing useful in that location.

These are worse when you're allowed to "recall from notes". Because your memories will converge on your notes. You add nothing that submitting your notes or reading them would.

And it's even worse when you watch the news, when you talk with others who were there, when your memories are contaminated with what others say they remember. Suddenly you find yourself correcting your prior testimony in the light of stuff you didn't have any memory of but now do.

Hanlon's razor rules.

Not that mistakes don't happen but here the tech just didn't appear to be good enough. It's a blessing they held on to the evidence all these years and the tech got good enough to exonerate this man.