Sean and Rich and Dan explained it well, as far as I know. Here is some more that might help.
The "fixable" problem via PCM flash has to do with the "primary" or upstream O2 sensor. As I understand federal emissions regs, a big change over recent years has been start-up emissions. Until the O2 sensor gets hot it won't read right and the engine can't go into "closed loop." Closed loop operation uses the O2 sensor output to fine-trim the injector pulsewidths to run the engine lean enough for minimum emissions. Anyway, the O2 sensor heater helps things along AND allows the computer to check the O2 sensor temperature, because the heater wire resistance increases with temperature. The PCM therefore supplies voltage & current to the primary O2 sensor and also reads the voltage across the heater circuit to see how hot the O2 sensor has gotten. If the O2 sensor does not heat up quickly (something like only 20 sec after start up!) then a code is set. I think the reflash basically tells the computer to give the O2 sensor more time to warm up. Along those lines, I also believe the '98-up headers were the result of emissions, as they are light and "double-walled" so that the inside tubing gets hot quickly to allow rapid transfer of heat to the O2 and to "light off the cat."
Right so far Rich, Sean, Dan et al?
Now, the non-fixable problem (as far as dealers and federal law for street-driven cars is concerned) is the downstream or "secondary" O2 sensors. OBD2 requirements added a secondary O2 sensor to ALL cars, such that there is one O2 sensor (primary) in front of each catalytic converter and one behind or downstream (secondary). A '96-up Viper and any other dual-exhaust car therefore has 4 O2 sensors, 2 primary and two secondary on the left and right side of the exhaust system. As mentioned above, the primary is used to directly control the engine in closed loop mode and is a good thing. It should never be tinkered with on a street-driven car, as it avoids overly rich or lean conditions in routine driving and allows the car to compensate for long-term changes in injector flow rates and other things. At WOT (wide-open throttle) the car no longer runs open loop, so the primary O2 sensor is also not an impediment to performance. So leave the primary O2 in place for all applications!
As far as I know, the secondary O2 sensor does only one task. It checks the "efficiency" of the catalytic converter. The whole concept of OBD is "on-board diagnostics," hence the ability to check if any component, including emissions controls, is not functioning properly (or absent). Mods just got more complicated, since you now have to know the strategy by which the PCM checks emission-related parts if you want to change them.
How am I doing so far?
I think (and this is where my limited knowledge starts to really taper off) that the secondary O2 is generally designed to compare its output to that of the upstream O2 sensor. For example, the primary O2 will "dither" back & forth around an output of about .5 volt. Higher voltages are rich, lower voltages are lean, and the computer constantly goes back & forth a little bit to average out right at stoichiometric (i.e. A/F ratio of 14.7:1). The cat then processes the overly rich exhaust by oxidizing the excess hydrocarbons and the secondary O2 will also dither, just not as frequently as the primary. I am sure there is engineering data that shows for primary O2 sensors that dither around say 20 times in 10 seconds, an "efficient" cat will result in a secondary O2 dither rate of about 10 times in 10 seconds or some such.
Hence the trickery. You can make a secondary O2 "emulator" (there are several on the market and some circuit designs out there) that will "dither" around .5 volts at some fixed rate that is hopefully within the range the PCM is expecting to see. Unfortunately, that may not be good enough. The PCM may also want to see a steady "rich" voltage at WOT and a steady "lean" when coasting down on lift throttle or may actually compare the dither rates or myriad other stuff. I suspect the PCM strategy may also have changed over the years, as the EPA probably mandates strategies that are difficult to bypass.
Now I'll bet that Rich and others may know more than they can say. It is very illegal to eliminate or bypass emissions controls and so they can't facilitate such efforts. So you can forget any dealer re-flash to disable secondary O2 sensors if you remove the cat.
For off-road purposes, I am also experimenting with secondary O2 sensor emulators, but have made precious little progress, since it is a pure hobby for me. I will post and crow over any success, but ONLY for off-road.
So for right now, it appears that you can add headers and fix any resulting check engine light. But if you want to remove cats, I don't know if anyone has successfully emulated the Viper secondary O2s. If they have, I'll take a set. You are likely better off with high-flow cats. That said, in the GM context, I have heard that high-flow cats tend to have a short service life and on my supercharged C4 I lose quite a bit of power with cats.
Geek-nerd signing off for now.