I think I'd go with a misfire in this case as well.

Like you said, all the ignition components could be called into doubt.
Replace the plugs, cap, rotor, and wires. Even wires that look perfectly fine can be trouble.

Saw 4 cars lately with wire troubles: a 94 Golf that had 2 burnt ends (looked like they rusted away at the plug end)
my 90 745, which simply failed one end completely due to 10 yr old wires
my friend's 82 245T, which was just one of a number of problems it had, but contributed to a misfire and idle problem- 2 wire ends pulled right out, a 3rd was starting to and we saved it temporarily
an 87 244 at the shop, with one plug end burnt away deep inside the connector.

It's more common than I ever thought, and often overlooked. That air inlet hose too, is only $15 or so at FCP, or $1 at a junkyard, so you may just want to swap a good one in to rule that source out.
--
Rob Bareiss, New London CT ::: 87 244DL, 88 245DL, 90 745GL, 84 242DL project, 89 244 parts, 88 244DL to replace the 87

Plugs are nearly twice the gap they should be. HC (unburned) is from a misfire...fuel went in, but never oxidized. NOx is generally temperature related but the excess oxygen (not consumed during a misfire) may be preventing the catalyst from getting into reduction range (excess CO2) required to reduce NOx. That the emissions are highest under load and drop well into normal just before gearchange indicates weak ignition. Use standard type, not platinum, spark plugs gapped no larger than .035" (0.030" is better). After doing that, drive the car a couple of miles and check the O2 sensor output for rapidly changing voltage. If it appears to take a second or two to swing back and forth, the sensor is sluggish - probably from the excess fuel - and needs replacement before you retest.


Don't waste your time polishing the rotor and cap with sandpaper. It may look prettier but it does very little in reducing jumpover voltage.

Looks like a slow/failing O2 sensor as the excursions appear to corrolate with load/speed changes.

One other thought would be a leak in the intake pipe between the AMM and the TB that opens / closes based on how the engine is leaning in the mounts. Had a Toyota do this to me a few years back.

jorrell