If the problem is due to adhesive failure, why would some displays not be affected? If we look at other functions on these computers that degrade, such as rubber parts that go all melted and sticky, it happens to just about all of them. While most 170s and 180s have the problem, many (including my own 170) do not. My PowerBook 170 has zero tunneling as far as I can tell, I’ve left it on for well over an hour with no signs whatsoever.
Perhaps moisture is what causes the seal to degrade? One interesting thing about my 170 is that I do not think it was ever stored in a musty basement. I have found that most of the old computers I’ve bought smell of old musty basement, which would suggest that that is where they were stored. The 170 on the other hand, actually smelled rather nice when it arrived. My guess is that it was stored in a closet or something, inside of its bag which was included when I got it. This works in favor of the moisture theory, but not the glue theory. That is all that I can give really, it’s my only laptop with an active matrix mono/grayscale display.
Yeah, my PowerBook 180 was also very well kept. Other than the missing back door, which was actually in the advert but the seller misplaced it (and I believe the hinges were actually good but for some reason it popped off) the unit is pristine. There are no scratches on it, the SCSI hard disk is whisper quiet, and the floppy sounds new. I would group observations like that under anecdotal...well, they can be useful observations but become anecdotal if we draw immediate conclusions solely from them. I think collecting observational data like how quickly does screen fail, how does it fail (always on left for everyone or does it differ), what trace does it leave after power off and how quickly does it recover, and experimentational of how outside temperature and humidity impact the screen (does it fail quicker in hotter temps, or is is just humidity), does running it in cooler environment change things (i.e. outside in the winter or in a refrigerator as I suggested -- I'LL RUN THE REFRIGERATOR TEST IS SOMEONE SENDS ME A BROKEN LCD PANEL). So that type of information could be very useful if collected.
With regard to some failing and others not, I've read that the back-light in some is more yellow vs in others it is more white. So again, part of the data collection would be % of failed with yellow vs white back-light. Also, when and where are the screens made (I need to look at the circuit board to see if it identifies date and origin). So there are likely different production runs and different facilities that created them. So your display may have been made in Taiwan whereas mine could have been made in the Philipines (I made these two places up to make a point). Local supplies of a sealant could have varied. I would also think that time and heat are the enemy of most sealants, causing drying and cracking. But there are a lot of PowerBook users with TFT screens that have the tunnel vision issue, so I would say those that don't either had there's made in a limited production run (maybe towards the end of the PowerBook model) or, as you said, some environmental condition that needs to be pinned down. But collecting data is better than drawing quick conclusions, that's all I'm saying. I fear that there are a bunch of folks stuffing their poor PowerBooks in bags with rice hoping things will improve.
So I think if we, as a community, want to get at the root of the problem, this type of input would be helpful. But to be honest, when I researched this stuff, the frustrating thing is I only read mostly anecdotal stories, like some guy who lives in Arizona not having any issues and therefore it must be moisture because it's dry there, etc... Reminds me of the Monty Python Witch logic skit: