Maybe some of you could help me out with this:

KE Jetronic, 124 230E 92, 2.3L
It had low CO readings and HCs were a little high, idle was reasonably smooth but had some vibration when warming up. At that moment the emission readings were:
HC 74 ppm
CO 0,02%

Then, I fixed some small vacuum leaks (idle hoses), changed injectors with seals and O2 sensor (90 kmiles). After that I learned (here) how to measure and adjust the duty cycle:
With battery voltage of 14 volts, voltimeter (at X11 / pins 3 & 2) gave 9 volts (average) at idle, thus the duty cycle was:
(1 - 9 / 14 ) x 100 = 36%.
Turned the mixture screw CCW (leaning), until reading diminished to 7 volts, so duty cycle at idle is now:
(1 - 7 / 14 ) x 100 = 50%.

I did an emission test with the new situation. These are the new numbers at idle:
HC 16 ppm
CO 0,001% (!!, is percent, already multiplied by 100, and it is 20 times lower than before).

Searching I found out that for this car (engine 102) the CO content of the emissions should be in this range:
0,5% - 1,0%.

Considering this, a 0,001% of CO though is good for pollution, it is far from how the system was designed. In addition, idle becomes smoother but is still a bit shaky when T° is raising and near 60°C (I suspect when the O2 sensor activates and the ECU's mixture corrections begins).

Could all this be indicating that there is still a small vacuum leak?
Please give me some light about what's going on here.

Thanks in advance for any idea.

Nothing is wrong!

If you review the service literature you will find that the CO specification for catalyst-equipped KE engines is less than 0.5 percent with no lower limit.

You did a good job tuning your car, and a little bit of "idle roughness" above 60C coolant temperature (when the system goes into closed loop and maintains an average stoichiometric mixture except at WOT) is normal.

Many of us in North America would love to see such low emission numbers on our M103 six-cylinders, but even in good tune, they are often in the range of your first test or somewhat higher, and most HC limits are a little over 100 PPM, so it's easy to bust the test, especially if the catalyst is not as hot as possible.

Duke

You are so right. The CO specification that I mentioned is for non-catalyzed cars (I got it from a book from the eighties).

The injection system is doing its work properly. It's just that after months of learning and trying I didn't realize that, finally, had arrived where I wanted.

Thank you, it's nice to know it.