It is sometimes difficult to determine if the switch is working, and nothing in the software tells you except the reaction of the various gauges like AFR's and Fuel Adj. The problem is that it isn't practical to work the throttle more than about 2-3% while idling in neutral, so if there isn't any or much difference between the open- and closed-loop values you won't see enough change to tell you anything. Since 2% TP @ 1500 rpms is easy to reproduce in neutral, temporarily change your Target AFR to something like 14.8 at that spot. This is assuming your base map is rich enough to show a difference when you switch. If so, connect the bike to the computer and launch the PC software, then run the engine at that TP and RPM. After the O2 sensors warm-up (about 1 min.) and you see AFR's show up in the software, flip the switch and you should see a change in AFR's and Fuel Adj. gauges in the software.
Another way is to put a passenger on the back with the laptop computer and have him or her watch as you (or the passenger) switch back and forth. That way you can access most any RPM and TP within reason.
Yes, I'm getting that, but mostly at 2750 and 3000 rpms. Go back to my original post in this thread and look at my sample trim table, but keep in mind I'm using 14.2 in that range. I'm not seeing it getting leaner and leaner like you are, but I'm not accepting trims anymore since it works against what I'm describing here. I really don't think this RPM/TP area is anything that would affect the bike's operation, as I think you'd run into it only on decel. Where is the weakness felt?
Where I did have a problem was high negative numbers (-25 or greater) at 1250 RPM's @ 2-10% TP and it was causing it to feel weak right off-idle. Here it seemed to get leaner in time, so I changed these values to "0". That tells AT to run only the base map values there, and that fixed the problem. You can see this in the sample Target AFR table in the original post. Since all of these issues occur at low TP, I tend to attribute this to reversion where air backtracks to the O2 sensor and forces erroneous readings.
Be advised that if you do this the trims will still be there and the PCV will still use them. Thus, if it was too lean before it will stay that way after you add the zeros, as the trim values won't go away. You'll have to Clear or Accept Trims, which zeros the trim tables, and then change the Target AFR values to "0" in those locations.