We are treading on shaky ground here. I'm going off what I've learned from years of the '90's and 00's Chrysler Mini-Vans and regular Cars.
Jeeps can be a bastardized system, some new chrysler, some old AMC, some Renault. As well, correct me if I'm wrong but most of the 80's Jeeps were Renix systems from Renault (French). And terms, techniques and implementation can be very different, especially between countries.
From what I understand, the "designed to run a little rich" depends on interpretation. Ideally, the most efficient would be to run even leaner than 14.7:1 at cruise and low load, but that would require a very expensive Wide-Band O2 sensor and also create a lot more NOx emmisssions. SO, in the U.S. regulations and manufacturer costs all lean toward running the stoichiometric 14.7:1 even though it produces lower mileage. So you can argue that American Cars run a little richer than other countries, but they are shooting for the 14.7:1 and hitting it accurately, so I, and most, would NOT say they are "designed to run a little rich" most of the time. Thats most of the time, remember the engine computer has an Open Loop mode.
Open Loop, the O2 sensor provides no Feedback, because the motor is in a mode where it needs a richer than 14.7:1 air fuel ratio (Warm-Up or WOT) or the O2 sensor is not warmed up yet and thus can't provide a feedback loop. In those cases, the engine computer is purely working off stored tables for how much is needed per rpm,air temp/press, engine temp, etc. The table is pre-set by the manufacturer, it was experimentally derived in a lab with a test engine. Since the system is adaptive, and no 2 engines are exactly alike, especially after they start to wear, air filter starts to get dirty, they are fitted with an aftermarket muffler, etc. The Engine computer can infer these difference from when its operating in closed loop mode with the O2 sensor. i.e. the engine computer gives it the amount of fuel called for in the preset table, the O2 sensor provides feedback that the preset table is causing lean condition, the engine computer adjust fuel to get back to 14.7:1 and then measures the difference and stores a corrective factor in another table. The table of corrective factors is also then used to when the computer is open loop mode, to compensate for changes in the engine. If you reset your PCM, the corrective table is in volatile memory, so it gets dumped. Then your engine computer has to start filling in a blank corrective factors table, and until it gets it filled in with good data, People will call this the "Learn Mode" while the table is filled, until its filled in with good data the A/F ratio will tend to be off a little more than desired, especially in open loop mode.
BUT, when in Open Loop, there is no feedback as to what the actual A/F ratio is at the moment. The engine computer can only make its best guess off what the parameters are, the amount of fuel it comes up will hopefully create the desired A/F ratio. Now, which is safer for a motor, going lean or rich? Going way lean, even for a few seconds can cause serious engine damage, especially at WOT. Going to rich, may create less than optimal power or emmissions, but you have to spend a very long time at extremely rich conditions to get the less damaging effects of deposits on valves, piston tops, exhaust, etc. So, when in doubt, biasing toward Rich is safer than lean, and when your in open loop mode, you are in doubt of what your actual A/F ratio is. BECAUSE OF THIS, IN OPEN LOOP MODE, MOST ENGINE COMPUTERS TEND TO RUN RICH. The Open Loop Mode is a minority of the time, Closed Loop is where your engine computer operates at 98% of the time. So its only fair to say, the system is designed to run a little rich in the open loop mode, where the computer is guessing at the A/F ratio, because running a little rich is safer for the motor than running a little lean.
Running Rich to get the Catalytic Converter to work right?? I've never heard that, don't know for sure. I was always under the impression that running stoichiometric was the best for the CAT, the amount of emissions being forced into the CAT determined how hot it ran. I.e. more chemical reactions, the hotter it ran, so more emissions would create more chemical reactions and thsu more heat. Running stoichiometric usually gets the best balance of emissions. Running to lean will create temp/press spikes in the combustion chamber, which creates NOx (also creates detonation and engine damage). Running to rich creates unburnt hyrdrocarbons and CO. The CAT does have to reach a certain temp to get the most effect, some cars do have heated CAT's to do this. But, I "think" the emissions will drive up the temp of the CAT to where it needs to be, its just a matter of how much emissions get by the CAT after start up, until it gets it temp driven up, determines extra measures to heat it up, once the CAT is warmed up, no additional measure are necessary to keep it at proper temp.
Ecomike, your description of how two O2 sensors work together sounds like it could work. BUT, the problem is, running rich results in very little oxygen in the exhaust. Thats why running rich is so bad for a CAT, there isn't enough left over oxygen to complete all those chemical reactions.
Perhaps some motors are less efficient design and don't have complete combustion? That may result in a rich mixture with excess oxygen? In that case, your 2 O2 sensors working together might be a good system to control the engine.
This is why I say we are on shaky ground, some of us may be talking about what we've learned from Renix systems and some of us are talking about what we've learned from Chrysler system on their more modern engine designs. We may all be right, just talking about different implementations used on Jeeps over the years.
There are some people that still swear that the downstream O2 sensor on the modern Chrysler systems is used for some adaptive information and it is NOT purely just to check the CAT. BUT, there is no references for this, as well, there are observations and basic logic from what is know that disputes that. Most of the detailed description of how the system works is proprietary, so I've learned more from discussions on forums than from any official document I've ever found on the system.