Your test is usually called the "two-speed idle test" and is the simplest field emission test. HC and CO are measured at idle and 2500 RPM, no load. It is deemed adequate for areas that don't have significant air quality problems. Areas that don't meet ambient air quality standards in the USA are required to do a road load test, which allows the measurement of NOx since significant NOx is only produced when the engine is under load.

Your 2500 RPM numbers indicate all is well, but the idle mixture is very rich as indicated by the high CO.

M103s often show high emissions due to insufficient catalyst temperature, and the temperature required for the catalyst to promote reaction increases with age, but if the engine was tested at 2500, then immediately tested at idle, I don't think catalyst temperature is an issue in this case.

It's also my hunch that the surging you report is related to the emission problem.

Checking for vacuum leaks is a good starting point. Also check microswitch and throttle position switch function. Then you should check the duty cycle. If someone has diddled with the basic mixture screw, the Lambda system could be operating at the limit of its control authority.

Familiarize yourself with following link and the links contained therein.

http://www.peachparts.com/shopforum/showthread.php?p=833484#post833484

Search on "microswitch" and "TPS" if you're not familiar with those components. The above link also has links to the shop manual data that you can view or download.

One needs to understand the somewhat complicated K-Jetronic system and learn proper test and troubleshooting procedures.

There aren't any silver bullets. You need to do some serious research and educate yourself, but my hunch is that the problem is not serious and will not be expensive to repair if you do it yourself, but it may be time consuming.

Duke