Well, the mystery deepens.
I spent last Friday night and Saturday disassembling and reassembling the three calipers and the PCV and bleeding the entire system.
The only components I didn't take apart were the three master cylinders and the delay valve.
I used the guides put together by HS and jay-d.
The work gave me an excuse to buy a Mityvac bleeder kit that runs off of compressed air (new toy!!).
All 9 pistons were in excellent shape and moved well. No signs of leakage or seal degradation.
Once I disassembled the PCV, its decompression piston moved freely in the bore and the rubber seals showed no signs of wear. There was no way to access the cut piston.
It was easy enough to establish that the front and rear M/C's were working fine.
When I had the left caliper disassembled, the SMC piston moved freely in its bore and I saw no signs of leakage.
I also took advantage of situation and replaced the front rotors with a set of Galfer wave rotors I'd bought some time back.
All three sets of bake pads were pretty thin, so they also got replaced with the Galfer pads purchased with the rotors.
I don't have a 42mm socket yet, so the rear rotor will have to wait.
Maybe I was doing it wrong, but the vacuum bleeder did nothing for me when trying to bleed the linked portion of the system (RMC to right center/RMC to left center/RMC to SMC/SMC to PCV/PCV to rear center).
Without someone to pump the rear lever, there was no flow.
My house-mate helped me on Sunday by pumping the rear brake lever while I bled the various ports tied to the linked portion of the system.
The one part I struggled with is bleeding the secondary master cylinder on the left caliper assembly.
Although it moved freely when I had the caliper disassembled, I never got the impression that the piston moved at all when I would open the bleeder and push on the piston bracket.
Low speed test rides around the neighborhood on Saturday indicated that the front and rear braking worked overall.
No binding occurred, even when I would draw down hard on both the front brake lever and the the rear pedal.
A 25 mile ride on Sunday, prior to bleeding the system with my house-mate's help, was also without issue.
I made a conscious effort to avoid using the rear pedal and didn't encounter any situation that required abrupt braking.
Riding to and from work today, however, was a different story.
I again made a conscious effort to avoid the use of the brake pedal.
Frequent braking at highway speeds, even though it was not particularly abrupt, was sufficient to trigger the problem.
I had to stop once on the way to work and twice on the way home to open up the center bleeders on the front calipers and release the pressure.
That was sufficient to free up the wheels.
I wasn't in a position to determine whether or not both wheels were bound, but I suspect they were.
The manual's technical features section on the LBS states that the front center pistons are actuated by the rear pedal; first the left front and then, if pedal pressure increases, the right caliper.
The rear center piston is actuated if sufficient rotational torque is present at the front wheel to cause the SMC to apply pressure via the PCV.
So while the mechanism to actuate the center pistons differs from front to back, the same hydraulic system services all three pistons.
Since I didn't use the rear pedal when riding to and from work, I don't think the delay valve is suspect.
My next step is to swap out the PCV from my other 5th gen to see if that has any effect.
I just realized, however, that I needed to open the PCV's bleeder when trying to depress the SMC piston. I was opening the left caliper's center bleeder. Doh!! :wacko:
I'll have to try again tomorrow night.
As always, any thoughts are welcome.