I know my system is totally different to the one you guys are using, in many aspects, so this may/may not be of any use .......
I had a similar problem some days ago night (morning ?? it was 0130), the OAT was -7C ... obviously the changeover took a tad longer than normal, but only by about 1/2 a mile, however instead of being almost un-noticeable there was a definite "cough" feel to the changeover, which lasted a few seconds, and a distinct smell of gas .... enough to cause concern ..... no lights or warnings anywhere ... smell went after about 3 minutes. No problems later that day in changeover/smells whatever. I checked the whole pipework with soapy water later .. no leaks found. Put it down to "one of those things".
Last saturday also had to drive in very cold .. -6C this time and the same thing happened ..... again no signs of any problems .....
Examining the system it occured to me I might know the cause, and I put it as my theory ......
on my system the vapouriser is down by the horns .. front left of car, with a significant pipe run from the vapouriser to the injectors. The vapouriser gets heated, the pressure rises, the system switches to gas, but the pipework from the vapouriser to the injectors is still significantly cold and the gas pressure drops for a short while, this causes insufficient gas at the combustion chamber and so misfires occur, the smell is the unburnt gas from those misfires ... as the vapouriser continues to work the pressure catches up with the normal requirement in a couple of seconds and everything works as normal....
Now, that is my theory, and I'm more than happy for it to be shot down .. but it might well be close to correct, and some elements could transfer across to your systems, as in slow to rise pressure due to abnormally cold pipework from the vapouriser to the injectors for a short time on changeover, especially as I believe your changeover is based on TIME and not on PRESSURE as mine is ??
Tin hat fitted awaiting incoming !!!
