I wanted to log further developments with my three N900s and this issue. The case study of N900 #2 is, I think, most useful/informative.

#2 continued to have the issue after I last mentioned it, but something had clearly gone very wrong internally - instead of just having the telephony-disabled error, it began to noticeably overheat and (seemingly) drain battery very quickly. I did not have time to fix it so I eventually left it turned off until I did. Two days ago I finally opened it up: turns out, the springs I put in put too much pressure on the EM shield. The EM shield got partially popped off inside there on one edge. Using my limited physics and electronics knowledge, I conjecture that the exposed chips were getting bathed in the EM emissions from one or more of the antennas, which in turn excited the electrons in the circuitry inside, which in turn generated heat galore. It's also possible the process caused some sort of feedback loop - the electric current generated in the chips causing connected hardware to recieve signals that mapped to instructions to do something, like maybe broadcast more **** out of the antenna(s). Another possibility, one of the springs somehow managed to slip into a position where it closed a circuit between spots that weren't meant to be connected, and the heat came from the resultant 'undefined behavior'. Frak knows.

Moral of the story: clearly, there is some 'runtime' risk to these fixes, if you mess them up. After having corrected the problem yesterday N900 #2 seemed able to use telephony again, but (given my very limited knowledge) it's perfectly conceivable that permanent damage to the circuitry thus affected could have occurred.

#3 held (after the addition of extra springs) for about two weeks. It then returned to having the telephony problems. I stuck a piece of paper at the top of the simcard (inside the sim holder itself). Worked for another one day or so. I stuck some more paper (two layers) on top of the sim card slot, held in placed by the battery. Worked for another one-or-two-or-so days. Replaced the outer paper with a folded piece of thick paper (torn off the outer packaging of a bubble gum pack). Worked another day or so. I tripped the layers of the thick paper - the back cover bulged even. But it worked again. Reduced the layers to just twice the first amount instead of thrice within a day, to get rid of the bulge on back cover. Has been working for a couple of days now without issue.

I document this so that people know - sometimes you need more springs, sometimes you need more paper, sometimes you need both, and gradually more and more of it. I also want there to be a proper chain of documentation on these 'hackish' fixes to this issue. I suspect many people fix their N900s this way, post a confirmation, then it breaks again later and they don't bother to inform the community of that part. So far, my assessment remains: these fixes are good for triage, but a routinely used N900 may fail again given enough time with just these fixes. Or it may not, if you're lucky, but currently there's insufficient evidence to determine with any real certainty one way or another.

I REALLY want this stuff to hold my N900s over until I can get a Neo900. But I also hate false hope (or lack thereof), thus incessantly logging the repeated failures and re-fixes here.