Klein Quantum II Ultegra shifting fix

Problem: The 3rd smallest cog presented problems while shifting. I could not reliably shift up or down to it. I could adjust the system every which way; but it was not possible to make it shift both up and down smoothly.
This problem is slightly worse for the outer ring. I've had a 42, 50 and 53t outer ring on this bike. I've changed the cogs thru 7 spd Ultegra to 8 spd ones. I've changed the shifter as well and the problem persists. I've not changed the derailleur but have no reason to think that it is at fault. I've adjusted the B-tension screw as far as possible. ie removed it.

After having gone to 4 different shops via 6 visits and spending hours with this bike myself the problem was still present. I'd declined an offer to send the bike back to Klein due to the hellish times I've had dealing with Canadian Customs. ie I would probably have to end up paying duties, handling and brokerage fees all over again! Let's just say that I was not happy with the shop that sold me this bike and have not delt with them since their two attempts to fix the bike.
Here is a close up picture of my fix. I basically re-routed the shifter cable so that it rotated the derailleur counter clockwise. This was done by routing it between the wheel axle and the derailleur and tying it in place with a ty-wrap or piece of wire.
This results in more chain wrap (5 deg extra rotation of the derailleur) and greatly improved shifting. My theory is that the Quantum I had the top derailleur pulley closer to the cogs than the Quantum II setup. One of the main reasons why the older Campy indexing didn't work at all was because they had too much space between the cogs and the top pulley so that the chain flex prevented the derailleur from forcing a shift. Personally, I hate Campy with a passion; it's over-priced, over-polished and under-engineered. It took them years to adopt the slant parallelagram [to their credit they did have one with an adjustable slant so that it could better adapt to a variety of cog ranges] and get any sort of indexing system actually working. On the flip side I'm not partial to Shimano constantly changing things. For example I see no use for 8 cogs; esp. a 12 tooth one. I rarely use a 13t and I'm never going to use a 12t.
With this setup I still get hesitation and often have to nudge it to get it to shift from the 3rd into the 4th and 5th cogs. I've set it up so that it shifts to the smaller cogs fairly reliably since I can't nudge shifts in that direction. When shifting to larger cogs I simply nudge the system by partially pushing the shift lever enough to slightly pull the cable; but not enough to cause it to lock the ratchet for the next gear.

The bike shops had replaced the cables, torn out all of the internal cable guides and the shifting was not in the least bit improved. I've been an amateur team mechanic and cyclist for many years; building bikes for friends and was similarily stumped after hours of adjusting the setup. Only after noticing that the bike would shift if I pull/pushed the derailleur closer to the cogs did I come across this solution. With my bike on a stand I can advance the shift leaver and the bike will not shift; but as I rotate the derailleur up it will shift. Anything that positions the derailleur closer to the cogs reduces the problem. Since a close tracking of the top pulley to the cogs is necessary for the indexing to work my theory is that the pulley is not close enough. Still, I can't image why, if that was the problem, only one gear is affected.

My uninformed guess is that the dropout change from the Quantum I to the Quantum II is responsible since it probably placed the derailleur further forward and/or lower thereby putting the top pulley further from the cogs. The Quantum I has a normal dropout while the Quantum II went to a "micro" dropout for reduced weight.
But that is just speculation. My Quantum I was T-boned by a car on the right side and I don't have any pictures of the dropout and derailleur position. Yes, I've been in a few accidents and missed quite a few. Take a look at my Automorons web page for the tales.