Ok, so i finally had some time to do some experiments - because driving around with no brake lights or brake lights constantly on is not cool. In short, i was eventually able to replicate the problem/induce it.
The system is more complex than one would expect, so i don't understand why Nissan chooses to do it this way on all their cars. My gf's Mazda2 makes do with just one switch assembly which activates the brake lights, cuts the cruise control and frees the gear selector.
If you press the brake pedal while accelerating, you will NOT be able to open the throttle any further than the position it was in the moment you pressed the brake pedal. When you let go of the brake pedal, the motor will roar back to life and accelerate as it should. Sort of feels like brake boosting in a way. This is supposed to be a safety feature, preventing you from accelerating while braking.
The brake lamp switch is a normally CLOSED switch. Ie. continuity only exists when the pedal is pressed.
The ASCD (cruise control) switch is a normally OPEN switch. ie. no continuity when the pedal is pressed.
These two switches act like an interlock, as described on page EC-505 of the FSM.
Therefore, disconnecting both switches or one at a time will not induce the problem.
To induce the super-lag, i had to simulate a depressed pedal. I did not want the brakes to interfere with my experiment as we all know the brakes can easily slow down the car. I already knew that pressing the brake pedal even lightly while accelerating will not allow me to go any faster, even with my foot to the floor.
To simulate a depressed pedal, i disconnected the ASCD switch and used a jumper wire across the terminals of the brake light switch connector (loom side). This is the point where i blew 2 fuses because the jumper wire shorted to ground. OOPS! Be careful
With this set up, i was able to induce the lag and learn a little more.
I could not get the throttle to open up more than 18 degrees. Test conditions were similar to the night i created the logs.
It seems that once the ECU realises you've been trying to accelerate while braking for a certain period of time (or distance?), it will switch to this limp mode where the throttle plates open extremely slowly and only to a narrow angle/range. I realised this when i removed the jumper wire and reconnected the brake light switch. Brake lights worked as they should (when jumpered, the brake lights were constantly on), but i could not escape this limp mode. The ECU at this point must be thinking the car is in a runaway situation. This explains why the problem vanishes after the engine is shut down and restarted.
Now, as i've outlined above, having one faulty switch shouldn't cause the problem. Bizarre? I think so.
Being electronic, i wouldn't rule out both switches failing at the same time as they are used equally as often as each other. My theory is that the rubber blobs on the pedal assembly could be worn out or causing the switches to stick in their respective pedal down positions or some gunk has accumulated inside the switches.