Thanks for the response. I checked the O2 sensor output a couple of years ago, and it appeared normal. Since my car is a CA model it has some limited OBD including O2 sensor health, and I have no check engine light or codes.
I've been reviewing the archives and cat temp appears to be an issue, as it is on some other cars. For my test two years ago I had it purposely done on a rainy day as CA has a "tire dry test" where they run on the dyno at moderate speed with no load for about 30-60 seconds before the test. I specifically asked the tech to run the tire drying protocol, and he did. This might have helped. The first test is at 15 MPH and this is where the HC is marginal. No problem on the 25 MPH test, and the 15 MPH test probably heats up the converter as it is done at a higher load than the 25 MPH test. This year I went on a dry day, but the car had been operating for about 30 minutes including a brief freeway jaunt, and the engine was not shut down prior to running the test. The coolant temp was at 100C when he began the test after doing the visual with the engine idling.
Car is delivering normal mileage - 20-22 in urban/suburban driving and about 27-28 at 75-80 MPH sustained freeway trips.
My archive search supports the suppostion that the M103 engine family seems to be marginal on emissions performance, but I couldn't find a silver bullet. Anyone else have any ideas?