Steve,
well, I don't think the people say "it works here" are doing it to annoy you. I think they are doing it to help give some idea about how many people don't have your issues. I think the greater number of people don't have the issues you and ano have. I know ano had these issues a while ago, and seems bitter about it. I would be too, as back then, "cheaper" UPB switches were $80+. I have 30 switches, $2400 in switches alone! + Face plates. I would have been bitter as well, especially since all the "reasonable" solutions do not work. I do not know, but ano probably couldn't just return the switches either.
I also have read (somewhere on these forums) that someone had a pool pump that killed his UPB. I've not seen anything do that to me. The same stuff that used to kill my X10 doesn't do anything to me anymore. Of course, I don't have a pool...and I don't have my hot tub yet...maybe one of these will render my system dead.
Personally, for my wiring, in my area, with my house, my electronics, "it works here". I don't even need a coupler and I can get signal strengths of 80ish on the same phase and 30 on the opposite. With a passive coupler, I get 80 on same and mid 60's on the opposite. I use the coupler as it does have some TVSS capabilities...and why not, it was in a kit I bought.
Again, personally, I can understand why SAI would say what they say...some people will never agree to it. That's up to how you read and understand what they say. My understanding is to make GenII "better", they made them hardware-deafer (stronger filters) and better software error rejection algorithms. By doing that they decrease the "distance" they can hear, but it allows them to "see" more precisely what they want to. So, in a "clean" environment, i.e. apples to apples...GenI will work better. In a terrible environment, depending on the type of noise, GenII is probably better. I say depending, as Gen1 can (my understanding...correct me if you have better/more current information) transmit more powerfully, and receive more "spectrum". So, if the noise is within the data, Gen1 can possibly transmit over top of it. GenII hears less, but hears more of just the UPB signal. That is, the noise has to be closer to RIGHT ON frequency to the UPB signal to effect it (i.e. folded into the UPB signal...so the extra filtering can't distinguish the difference). So, depending on what the noise is, the signal to noise measurement might show Gen2 better, might show Gen 1 better. Chances are, in a noisy environment, Gen2 will work better due to it's ability to reject the noise. Gen 1 can only try to overpower it. Keep in mind, they are using an RFPIC to read the data from the powerline...so signal to noise is critical...and will depend on what noise is there.
If I had an issue, I probably would have went Gen2. I got a kit. Put switches and modules on both phases. Tested it out. Everything was 100%. So, then I made my big purchase and loaded up the house. Haven't had an issue or looked back ever since.
If I had, had an issue, I probably would have purchased the web mountain parts. I love the way SAI switches feel, so going web mountain would have been my choice.
If $$ isn't that much of an issue (it was for me), then just go Gen 2. It may end up costing you (from memory, so I could be wrong on this number) $5 to $10 more per switch. It won't hurt, and depending on what noise MIGHT appear later...heck you might be better off. For me, I might get screwed because of my gen 1 stuff. Then again, (my feeling is) I might get struck by lightening...or hit by a car on the way home. It's all a risk and determining what you can tolerate. I do feel bad for Ano and tenhold, and I know they are both sharp guys...so to have purchased this stuff because "it just works" and to not have it work...it makes me think, maybe they should have went Z-Wave. I mean, knowing their history, I doubt it is something they did to cause this. It's really annoying...as I don't think there are too many cases of UPB not working, so how do you troubleshoot something that doesn't need troubleshooting (more times then not)? X10, everyone can help, and can offer suggestions --we've all used X10 at some point and ALL had things go crazy when something like a cellphone charger...charges. UPB...there's few people that I think have ever needed to troubleshoot...which leaves people like Ano and Tenholde flying in the wind.
For me, I've wanted to try (as an experiment, with my neighbor's consent) controlling an appliance module from my server, plugged into my neighbor's house. We're on the same transformer...so, I do not see that it won't work...but maybe...I've been told in good conditions, Gen1 should be able to be seen by a cluster of houses, as the signal is so strong. I've wanted to test this...
Wow, what a rambling mess...sorry Ano...maybe there is something in there that is useful to you? I personally think that ano's problem is related to his wiring...as I can't understand how you get such terrible cross phase signal strength...even with multiple couplers. Maybe your only solution is go Z-wave and accept Z-Wave's issues, or start swapping switches to Gen2, so you can get the Gen2 active coupler (which sounds like might be what you need to solve your issue).
HTH,
--Dan
P.S. Side thought, (and the reason for the edit), If you move the PIM from one phase to the other, do you "resolve" your issue for that phase...as in you are plugged into phase 1. Phase 2 has issues. Move the PIM to phase 2, phase 2 gets corrected, but you loose phase 1...if this is the case...at least for your situation...
what about putting one PIM on each power phase...then using xAP to transmit /rx on the appropriate phase? I know you don't use Homeseer any longer, but does your new control software support xAP?