I've been thinking about this "mod" for a while and here are my highly scientific, prolific, and you get what you pay for thoughts on what it does, why it exists, etc.;
I don't believe this is a hidden software engineer program workaround.
I believe this exists out of necessity due to the world we live in where lean is clean, or something like that. Emissions, emissions, emissions,.........
If anyone here has spent any amount of time tuning motors with the old carb set up and remembers what happens when the bike is lean jetted you will have an idea as to why this, IMO, exists. Lean conditions cause the rpms to "hang." This makes shifts balky at best.
So, if the clutch switch signals the ECU to go "rich" when engaged the shift points will have a good response from the motor during shifts.
You would think that one could tell a significant difference during neutral, non moving conditions when the clutch switch is engaged but I would almost bet that the algorithm requires an input from the speed sensor to initiate the "rich" condition.
The bit where the bike falls on itself sometimes just after take off and during clutch engagement? What if it is caused by the bike seeing movement and a clutch switch activated initially (during friction zone) so you are getting the "rich" condition. As the clutch lever is let out the switch disengages and then the ECU returns to the lean map. At that moment the bike would fall on itself and then pick back up under the new "lean" condition and stock mapping.
These are my thoughts on the matter. Thoughts, ideas, etc.?