After examining all charts I can find - stock, various mod combos, AND the reprogrammed ECM dyno runs from Best Dual Sport Bikes, I have to report the findings.
Indications all point in the same direction - a ECM running software that purposely "messes with" power production above 8k rpm.
I say "messes with" because the charts all show a weird "mangling" of the power curve - waxing and waning from 8k to about 9k rpm, then a smooth drop-off to redline.
This is commonly referred to as a "retarded ignition" detune. I can find no actual evidence/data that this is the case.

The subject of Ignition curve theory and development for the upper 1/4 RPM range of any ICE is more than a "Dark Art", it is a BLACK HOLE. All one can do is to research the subject to glean all available information, form a personal theory, and experiment. Basically, sometimes power can be found by creating some kind of shape of a retarding curve or some kind of shape of an advancing curve - over the stock curve.
With the carb'd KLX250, the ignition is set at 25 degrees for the entire RPM range ( other than idle ) which made theory execution, and testing of results, very easy. I tried many shapes in the upper RPM area ( advancing and retarding shapes) without any significant power increases over the stock flat 25 degrees.

There is only one way to solve a ECM detune program - flash it with a better program.
At this time, only Best Dual Sport Bikes is creating and flashing custom programming on the KLX300 ECM.
As per the BDSB website page for their KLX300 ECM, their programming includes a "restructuring" of the ignition curve in the upper RPM range.They claim to have RETARDED the ignition curve in this upper area from stock. This could very well be true as it does adhere to sound ignition theory for a upper end.
It seems their efforts - to date - have, in fact, created a better upper RPM power curve. Which also indicates that the stock programming is incorporating a detune of the upper end.

 

Hopefully, BDSB will continue to refine their KLX software as they need to defeat the secondary throttle valve actuation within the software and get rid of the remaining temporary dip in power in the upper end.

 

Since BDSB will NOT program for performance exhaust systems or a lidless airbox, their current ECM software can only be used as a better "baseline" for a truly high performance KLX300.
The mod combo would be as such: Secondary Throttle removal, Barker exhaust system (O2 sensor must NOT be deactivated) , lidless airbox, BDSB ECM, and finally a DJ PCFC with a trip to the best local dyno shop for a quick refueling of ONLY open loop operation to 12.6 - 12.8 ( inclusive) AFR.

Results will be well over 28 hp along with a huge TRQ curve.

 

BDSB dyno charts need to be taken with a grain of salt as the upper rpm range (even on stock bike) looks vey different from other dyno runs(dirtrider, JH motor club and number of this forum members including tooter).

 

Well hold on. I've fully analyzed his charting with the others. His charting is accurate. I would not post his stuff if it were not..
He is not using a DynoJet dyno. This means that the smoothing function of the data points( therefore curves) is quite different than what is produced with a DJ Dyno. His curves are heavily/overly smoothed. This means that only the "trends" of the power data is displayed as a curve - there is no granularity at all.
When this is taken into account, as well as his chart grid having an oddball scaling, his power curves do match all others.

However, none of this is of any importance to the subject at hand - (ECM programming that detunes the top end).
His charting of the stock power curve and the curve of his latest ECM program - All on the same dyno - proves that stock ECM programming is sub-par over 8k rpm.

 

Exactly! Due to high smoothing his charts don't show that huge drop after 8000. Other charts show much steeper drop.

 

I'm enjoying reading this conversation.
Is it possible that the intentional factory upper range detune is for engine longevity?

 

Unlikely to me. It would be easier to put hard rev limit if that's the risk. On top of the older klxs rev to 11k rpm without detune.

PS I have bought several klx250 JDM performance ECUs which I hope don't not have 7500rpm detune. I hope I will post results soon.

 

The granular charts show a very strange behavior 8k-9k rpm (approximately) .

Such odd data would normally indicate a malfunctioning component - but such seems to not be the case with the KLX300. However, I can think of no reason why the engineers would purposely "foul" this 1k rpm segment - so I had to think "outside the box" on it..

Could it be possible that they want to try to get peeps to shift between 8k-9k instead of 9k up. By providing a little temporary power slackening, they were hoping peeps sense it and would shift sooner ?

All I know is what I've posted, BDSB has significantly enhanced 8k and up power production thru revised programming.

 

Might it be a little bit of tire slippage?