It is possible to be a bit more scientific about all of this. The O2 sensor and the diagnostic port can supply a lot of info about rich vs lean operation if you make the right measurements.

Here are a couple of good links for explanations of how all of this stuff works and where to make the measurements.
http://www.autoshop101.com/forms/h25.pdf Pages 11-15 describe how the injector control works and how the O2 sensor and the catalytic converter work together to keep emissions down.
http://www.autoshop101.com/forms/h27.pdf Page 13 describes how to use the VF terminal on the diag connector to figure out lean vs. rich.

In a nutshell, here's how the system works.
- The ECU has a "map", or table, of injector times based on air flow, air temp, rpm, and coolant temperature (those are the main ones.) It uses this map to calculate optimal injector timing for engine conditions. It recalulates the injector times pretty much on every revolution of the engine. Generally, if the sensors are all working properly and the injectors are stock, the engine will run well with this information. It will not, however, necessarily pass emissions.

- To pass emissions, the cat converter has to be operating with just the right mixture of combustion products. It is designed to work well only if the mixture is at the 14.7:1 air:fuel ratio, the stoichiometric ratio. That's the air/fuel ratio where there is just the right amount of oxygen to combine with all of the fuel and burn it completely. Page 14 of the first document I referenced shows how fussy the cat is with respect to air/fuel mixture. The purpose of the O2 sensor is to get that mixture right so the cat can do its work. This only happens when the engine is in "closed loop" mode, which requires the O2 sensor be warmed up and the engine running above idle but not at full power. That's why they use 2500 rpm for the test, and nowadays put the vehicle on a dynomometer.

You can measure the VF terminal under those conditions in two ways. With the T (or T1) terminal open, the voltage will tell you if the ECU is trying to enrich the mixture (voltages near +5V), or trying to lean it out (voltages near 0V). Nominally the value should be near 2.5 volts. If it is not, something in the system is biasing the mixture away from nominal. These values are "learned" from the O2 sensor over several hours of operating time.

With the T1 terminal shorted to E1(ground), the VF terminal will track the O2 sensor output. This should be oscillating between 0 and 5 volts, at a rate slightly less than once/second . If it is stuck at 5V, the O2 sensor thinks the mixture is too rich and is telling the ECU to lean it out. If the voltage is stuck at 0V, the O2 sensor is telling the ECU the mixture is too lean, and needs enriching.

Gevo, since your smog test indicated a too rich condition, the O2 sensor output should be stuck at +5V. If it is stuck at 0V, the O2 sensor is suspect and is probably the cause of your problem. If it is oscillating properly, it means your cat is not doing its job even with a perfect mixture going into it (if the cat were healthy and the mixture going in was correct, you'd pass smog.)

If the O2 output is indeed at +5V, it and the cat are likely healthy. Something is causing the ECU to put too much fuel into the system. Among others, the following things will cause the ECU to improperly enrich the mixture: (essentially, all of these things cause the ECU to operate on the wrong part of the "map" which tells it how long to pulse the injectors.)
- Coolant temperature sensor giving an inaccurate, cold reading. Colder engines require richer mixtures to run well.
- Air temperature sensor giving a cold reading. Cold intake air requires a richer mixture because it doesn't vaporize the fuel as readily.
- AFM set wrong. If the ECU thinks more air is entering the engine than is actually happening, it will inject too much fuel.
- Injectors too large for the engine. The ECU determines proper injector timing based on a specific injector design. If that changes, it has no way of knowing it needs a shorter time. It can "learn" a new setting from the O2 sensor, but the range of learning is limited. Once it hits the limit, the engine is essentially in open loop mode all the time and will fail emissions.
- Vacuum leaks also distort the information the ECU receives, but normally that would cause a lean condition, so is likely not your problem.

I think if you measure the VF terminal both with T1 open and T1 shorted to E1, you'll know a lot more about where to look for your problem. Report back here if you want more help interpreting the results.

Good luck at getting past this hurdle!