Well, I can understand how MAYBE a clean install would fix the problem Perhaps some hardware driver update, protocol stack software, etc. got corrupted somewhere along the line of upgrade from 10.4 to 10.5 to 10.6, whatever. This would certainly explain how my MBP worked 'back in the day, and now is a POS. HOWEVER, this does NOT explain the numerous MPB users experiencing same problems with new MBP right out of the box from Apple.
Further, application level programs should never, Never, NEVER ever be cognizant of specifics down at the hardware driver level, or the network protocol stack for that matter. To do so would a a violation of some very basic tenants of software/system and networking architectures that have been around for well over 2 decades. If Apple were doing something remotely like this it would be a major no-no, stupid and amateur. I give Apple more credit than that. Thus, any notion that a clean install would fix the problem, then bringing applications and user docs/files would then somehow subsequently corrupt the WiFi networking simply does not hold up to scrutiny. I can't say it couldn't happen (memory leaks, applications corrupting system memory, etc notwithstanding - these being unlikely, again due to OS architectures and protections).
So, the only conclusion(s) I can make from all this:
1. There is a bug in the OS, hardware drivers or WiFi protocol stack (or any/all of the above) that has been introduced since 10.4 (last time I remember my MBP working solidly on WiFi), and seems to be specific to the MBP (otherwise explain side-by-side reports of MBP and iPad, iPhone, etc, with MPB failing, other devices working fine), OR
2. We're victims of mass hysteria. I can't vouch for everyone else on this forum who have reported ongoing problems, but I'm fairly certain I am not imagining the problems on my system, so I think the likelihood of mass hysteria is near zero.
What I do know for certain is:
1. I replaced my WiFi card in my MBP. with an airport extreme board. This improved my RSS levels somewhat, which is good. It did not improve my connectivity/stability much if at all.
2. I can sit in various locations in my home, not very far from the router, with VERY acceptable RSS levels, and the connection is flakey as ****. I can take airport down, up, down, up, and if I do this enough times I can get a connection with good RSS AND good bit rates (36 Mbps or more). Same location, all else the same. This is true in these locations all the while other WiFi devices, including iPhone and a WinDoze laptop working all day, every day, no hiccups.
3. When doing #2, I can get unstable connections from time to time, which start out with good-excellent RSS, and good bit rates, but within a few seconds or perhaps a minute or two degrades to 1 Mbps (which may actually be zero I suspect because I cannot, no way no how get email, send email or surf with Safari). If, on the other hand when doing #2 I get a good connection, say 36 Mbps by trying enough airport on/off/on cycles, it seems to work well for a good amount of time (but not forever).
4. I can sit right next to the router, get good-excellent RSS, get a stable connection which seems to go fast, last a long time and works with email and Safari with no hangs.
What I conclude from this is there MUST be a bug somewhere in software. In areas with somewhat reduced RSS levels (but still WELL within acceptable minimums), the software must be going stupid when it attempts to correct/retry errors, step down the bitrate, SOMETHING like that. I've written enough software, including networking protocol stacks over the years to know what a software bug looks like. I'll point out that my experience is NOT with Apple software or 802.11 specifics. Nevertheless, protocol stacks are protocol stacks, more or less. This problem sure as **** has all the symptoms of error recovery logic going stupid at some operating margin/boundary condition. This is typically where most bugs occur anyway, that's where I would be looking if I was Apple trying to fix this one. I also wouldn't rule out the possibility of an 802.11 protocol incompatibility with non-Apple Airport routers. Another source of software bugs - products that work well with mfgrs. other products, in the lab, all day long. Then they fail in the field in conditions not found in the lab.
Anyway this is my long-winded $0.02 worth. Maybe someone at Apple will take a look and it will trigger something that leads to a solution. I know for certain this is really, REALLY getting to be a PITA and I would appreciate a solution from Apple instead of this constant churning in the forums.