I don't understand your point about it having to suck air from both sides.

Looking at the diagram you posted (which is accurate enough for argument's sake), if there were a slug of air from the switch to the end of the block it would easily be trapped there during bleeding. Why? Because it's in a dead part of the circuit. If Joe and Dennis are gently bleeding the brakes as advised, there would be little pressure in the system to push the air out through the switch. Once the brakes are bled with the air slug still present, pressing on the brakes would let the pedal drop because the air would be pushed back out through the switch threads. Let go of the pedal and it gets sucked back in. Hence, you could bleed the system with little or no air coming out yet experience a soft pedal or no pedal.

I've had my share of weird brake problems and this is completely possible within the realm of my experience. I wasted over a week with a customer's car once because no one, not even the dealer, could tell me how to trouble shoot GM's new quick-take-up brakes (remember those-people still don't know that you can't manually bleed them). That was around '81 and that's when I bought my pressure bleeder. Year's later I had a similar problem with an '86 Grand National that I owned. After taking it to three Buick dealers, the last one (who employed a regional GN guru) wanted to replace the whole brake system and though it was under warranty, I wasn't about to trust the car to these guys. Using methodical troubleshooting techniques I determined it was a $49 accumulator that took 15 minutes to replace. Just because these systems are simple it doesn't mean that they are easy to troubleshoot and anything with moving parts or seals should be suspect when troubleshooting them. I don't think we would be belaboring this if it had been the front wheel cylinders or master cylinder.

If a small leak at the wheel cylinder could cause the symptoms that Joe experienced then why can't the same leak at the distribution block cause it?