So, I don't like how the pilots are getting all of the blame for this crash and feel that I need to get this off of my chest. I flew in to Buffalo that same night about three hours prior to Colgan 3407's arrival. We were in a B737-200 and coming in from Pittsburg. There was light to occasional moderate rhime ice between 5500 and 2500 feet. (Nothing that bad.) We were being vectored in from the South for the ILS 23 approach (sound familiar?). I was the FO for this leg and the ATIS had reported under the Notices to Airmen that "runway 23 glide slope information not available to the right of the approach course". So, I informed the pilot flying of the NOTAM as part of the approach brief. We were given a final vector and cleared for the ILS RWY 23 approach just outside the approach gate (7-10 miles) just like Colgan 3407 was. I am not sure if you know this, but the autopilot system is very old in the 737-200 and we are always used to it failing during a coupled approach. So, like clockwork, we were monitoring the flight controls closely. We were IMC and I happened to notice that the aircraft started a very gradual descent before we captured the Localizer. I mentioned to the PF that we were not aligned with the runway and that we were descending. With that, he said that he was going to hand fly the approach. Then, I noticed that we were momentarily above glide path, and then we went well below glide path within seconds. The PF noticed it too and said, "Hey, make sure we have the right frequency dialed in…" We both confirmed the frequency and before we knew it we had shot through the localizer. ATC called and asked if anything was wrong and I said that I thought we were having a problem with our autopilot. ATC's response was… "No, it's the glide slope; we have been having problems out there since they put in this temporary damn near the approach end of 23…" "No problem, guys, I will just vector you in from the other side." So, we got vectors back in from the North side of the field and the landing was uneventful. I can't help but to wonder if Colgan was coupled up like we were on the right side of the localizer and they got the same indications as us… If that was the case, then that would explain why the aircraft chased after the glide slope in the up direction. As you mentioned before, the Q400 does not have auto throttles, so the PF would have to be really on top of things in order to mash the thrust levers forward. I was reading the CVR transcripts too and it happened at the same place that it happened to us. I also noticed that the FO did not brief the CA of the NOTAM in the ATIS. I wonder if when she got ATIS information Romeo, if that NOTAM was still being broadcast. I think that the NOTAM should read "glide slope information erroneous to the right of localizer". If you look in the AIM 5-4-3 it talks about False Glide slopes. I believe that is what we experienced that night. I placed a phone call to the NTSB, but no one has since called me back. I, like you, will let the investigation play out, but I truly hate how no one has mentioned a faulty ILS system. Just for grins, and to cover my arse, I doubled checked my paper Notams (Local and FDC) and there was no mention of the glide slope having a problem before we launched to Buffalo. I don't believe that the crew should bear 100% of the blame, but like you said, I think that "sterile cockpit" techniques would have had them talking less and watching the instruments more. Either way, it's a shame to lose fellow pilots. PS: I heard that the pilots at Southwest have an internal NOTAM system where they mentioned the False Glide slope problem at KBUF. Not sure if that is true or not.

https:// fl250.blogspot.com/2009/05/thirty-seconds-of-silence.html