As stated in
Mech 1 twa's post, The PAIR system reduces emissions by providing a path for filtered air from the air box to be sucked into the exhaust port by the lower than ambient pressure at the trailing end of the exhaust gas pulse (after the exhaust valve closes). This fresh air mixes with the hot exhaust gasses; and at that any unburned hydrocarbons in the exhaust are burned--this does of course create heat and cause the exhaust valve, exhaust port and head pipe to run hotter than they would otherwise.
It is often called a "pulsed air injection" or similar, however that is a misnomer as it does not "inject" air; it just provides a path for the low pressure tail of the exhaust pulse to pull in air. The full path on most systems is as shown below.
This is the Suzuki Burgman 400 system. Suzi's service manuals are better than Honda's, with a lot more detail, and a lot of "why"and "how" information. The arrows and annotations are mine, the PAIR info is in green, red is the PCV flow.
You can see that (when the PAIR valve is open) filtered air from the airbox can flow through the solenoid valve and as far as the PAIR reed valve. Then when the exhaust pressure falls below ambient air is drawn into the exhaust port. Disabling the PAIR system causes the exhaust valve, exhaust port, and head pipe to run cooler.
The Silver Wing system works the same, and is actually physically very similar.
The ECU controls the opening of the PAIR solenoid valve as it sees fit using a combination of data such as ECT, IAT, throttle position, manifold pressure and engine speed. Typically it is open at idle and closed or near closed throttle deceleration.
Disabling the PAIR system was, on the '03 to '06 400 Burgmans, a common "fix" to prevent exhaust popping and full blown backfiring on deceleration.
With the '07+ S'Wing disabling the PAIR could cause the O
2 sensor to believe the mix is richer than intended. I am not intimate with the later model;s EFI system however if like many this would only be an issue at idle where closed loop operation might lean the mix if the PAIR is disabled. On closed or even partially closed throttle deceleration it would be quite odd (in my experience) for an EFI system to attempt to closely control AFR using O
2 sensor feedback.
It would be simple enough to unplug the valve connector and ride a while with a "weather eye", monitoring for ill-effects...