addyguy might be on to something. Hondas are tuned for efficiency, running as lean as safely possible.
So let's say your K&N does indeed "flow more air than a paper filter," you're only going to push it leaner on your Honda MAP-based EFI system.
There is no device that directly measures intake air mass on a Honda (AFAIK unless the newer ones are different). It's just assuming it's getting the amount of air it thinks it should be getting at any given time based on senor inputs and RPM.
The "strategy" of most modern ECUs is such that they will "adapt" and compensate for variations in A/F mixtures over the long term, to a certain percentage until a fault code gets stored and the MIL lights up. Like when a small vacuum leak slowly gets bigger over time, the ECU will compensate only so far.
BUT, short term fuel trims have very loose "tolerances" so to speak. Cruise along the highway with a really nice ($$$$) scantool hooked up to a modern car in recording mode at 65mph and blip the throttle. I've seen the instant/short term fuel trims spike to 15% (lean), or even higher for xxx milliseconds before the ECU can respond. ECU don't care. There is also a corresponding delay in the load based ignition maps. Double det. trouble, but a 16 bit ECU is typically worse than a 32 in that sense. Honda 16 or 32? Anybody know?
Typically it's no biggie on a well running engine, but you're doing that with "increased air flow" (maybe) on an already lean running EFI system that cannot directly measure airflow/mass.
On the flip side, a close friend of mine bought a 2001 civic coupe brand new and immediately installed a "hot air intake" with an open element K&N filter. That's it, no tuning at all. He drove the car up until just a couple months ago when the clutch went out, 145k on it. He sold the car for $2500, filter not cleaned or replaced ONCE its entire life, ran fine. True story.
At this point, if it were my car, I'd probably just stare at it for a little bit while smoking a cigarette, then forget about it.
