One of the big things you learn at uni, is to work methodical. Formulate a hypothesis, then conduct an experiment and ultimately compare the outcome with the hypothesis. Then take that as the basis for the next iteration of research. What sounds incredibly lame, becomes a lot more exciting once you throw in some real life at it...
All of this is incredibly easy, if you only have one variable, which if you're lucky you can even directly control. Let's say carb needles for example. The Mikuni manual for TM38-series carbs specifies 3 alternate needles for this carb. Each of them not only being richer than the previous one, but also differing in where the taper starts, starting diameter and final diameter. (For simplicity's sake they are all the same length.)
So, because the carb is running slightly lean at midrange, one swaps out the needle... and opens a can of worms of epic proportions. Because for example as now the initial taper is vastly different, the pilot jets are completely off...
All of this doesn't exactly get easier, with the wideband sensor deciding that this was its last hurrah and no matter what you do, it will only register 9.9 or 10.0 from here onwards, which is not quite as useful as it used to be.
Another method is to conduct multiple experiments and then compare the results. This is extremely useful, when you are VERY sure that you're on the right path and need a bit of dampening of your optimism. On other occasions it can be rather eye opening. For example I didn't notice anymore that my VM38s were stuttering at very small throttle openings... even after being freshly synced.
So I swapped back to the TM38 flatslides again, but this time with smaller 17.5 pilots and back to stock needles. Guess what, it runs leaner down low and the slight habit of fouling the plugs seems to be gone and oil-temps appear to have remained the same. I think I am pretty happy with the setup by now.
Undoubtedly though, the old girl looks the part.
Addendum: By now there's nearly 240km more on the clock and the average fuel consumption has gone done to 6.1L/100km.
Final jetting: 17.5, 185, stock 6FJ41 needle in mid (3rd) position, airscrew 1.75 turns out