Intermittent P0420 OBDII Code - What Should O2 Sensors Be Doing?

Have been troubleshooting the subject intermitttent OBDII code on my 97
Taurus 3.0 GL Vulcan with an AutoXRay reader.

While monitoring with the reader over the last 2 days, I have had the code
come in once or twice (low cat efficiency bank 1) and on subsequent scans
after resetting, its not there.  I have had this code come and go,
eventually resetting itself within OBD on its own, for about a year now.

In the real time data monitoring mode, both the Bank 1 O2 Sensor 1 and
Sensor 2 voltages are all over the map, between 0 mv and around 900 mv.

Same thing is true on the Bank 2 O2 sensor outputs.   However, I have never
gotten a low cat efficiency code from Bank 2.

From what I have read, the upstream sensor voltage should be varying, as
observed, while the downstream sensor voltage should be low and steady.
Having both vary would indiacte that cat efficiency was indeed low. If thats
true, why doesn't the Bank 2 fault code ever show up?

Does OBD II average the voltages of the two sensors 1 and 2 over a period of
time before it compares them to calculate cat efficiency?  I am wondering if
the P0420 code is being generated intermittently if the upstream Bank 1
sensor  goes to a low voltage, simultaneously when the downstream sensor is
at a slightly higher voltage.  Maybe the code is a result of a 'sluggish"
downstream sensor?

Should I be replacing the downstream sensor on Bank 1? Which bank is Bank 1
on the 3.0 in the Taurus, from the forward cyclinders, or the aft?
Defineitly want to be sure the sensors are Ok before I have to spring for
that big bucks ford Y pipe with both cats.   Car has 118K on it, otherwise
rnning well.  It may have a slight head gasket coolant leak, as I
periodically need to top up the expansion tank, with no evidence of external
leakage.  No indicaton of coolant in the oil or tranny fluid either.