A friend bought a T3 a while back that had been fitted with size 142 main jets, which the PO thought should be appropriate given the pod filters and Bub exhausts. (It was otherwise jetted as stock, and the motor had been properly rebuilt.) Since the bike smelled strongly of unburned gas when it went down the road, different sized main jets seemed worth a try. Both the stock 120s and larger 125s were considered. The 125s went in and the bike was magically transformed. It no longer required 4000 rpm to stop stumbling, and gas mileage went up from about 35 to well over 40 on a long test run.
My own T3 also came to me with 142 main jets, which I replaced over time with other sizes, ending up with stock 120s. My motor has good compression (above 170) and runs very well.
Seeing my friend's success with 125s I decided to try those again on my own T3. I remembered someone here citing Dave Richardson's recommendation for a 10% increase in jetting when using the K&N pods I have (and I found it on page 7-29 of my copy). My current K&N's are pretty new, so I figure they flow as well as they can.
By the way, I have English Feked stainless mufflers, with solid stainless baffles. I don't think they are likely to have a measurable effect on the T3 motor, given its mild tune.
Going from a 120 to a 125 jet is about an 8.5 percent increase in area and flow [(125/)120)^2 ~= 1.085], which is close to Dave's recommended 10 percent.
My seat-of-the-pants impression of the effect of going to 125s was disappointing. The bike seemed less responsive. But since people's seats are calibrated differently I decided to go the full science route in search of actual evidence.
My iPhone (and probably your phone too) has a built-in accelerometer, mostly to tell it which end is up. I calibrated mine on a kitchen table, observing that it produces the correct readings of 1 g straight down, and 0 g side-to-side. (The Bosch model used is listed in the Settings menu.) I used an app called Physics Toolbox Suite to record linear acceleration in meters per second-squared, in the forward-rearward axis of movement.
I have a test facility called the University of Wisconsin-Madison just down the street, with long, level stretches of road beside the soccer fields. My test procedure was to accelerate full-throttle in first gear from about 15 mph to 6000 rpm, recording the data. This seemed appropriate since the main jets should be the only carburetor settings influencing full-throttle acceleration (assuming the gap at the top of the aerator is big enough). 6000 rpm in first is a bit over 40 mph, so there wasn't too much to worry about in terms of law enforcement.
I did six runs with the 120 jets and seven with the 125s. Here are plots from the first runs of each series:
The plot on the left is a run with the 120 jets; 125s are on the right. A large amount of noise in the accelerometer output was smoothed away with a 60-period (0.6 second) moving average, shown by the overlaid dark lines. All of the #120 runs exceeded 4m/s^2, while none of the #125 runs did so.* The performance difference was very noticeable in the full-throttle runs: the 125s felt rough and stumbling compared to the smooth 120s. I began to think something was going wrong with my motor during the 125 runs, and was surprised that the graphs were all similar when I got home.
Sometimes carburetor problems really are carburetor problems, often self-inflicted. Guzzi T3s are free-revving, fast-for-their-time motorcycles capable of 115 mph. If you have one that can't get beyond 90, maybe your main jets are too big.
That said, there are many other things to think about with respect to performance in general and carbs in particular.
I hope my test results** and the reported results with my friend's T3 encourage someone else to think twice about main jets.
------------------------------
*A single #120 run was dropped because it showed much less acceleration than any other graph, due to rider error I suppose.
**The data collection and graphing were all done on my iPhone using the free Physics Toolbox Suite and the free Apple Numbers app. Both apps have capricious interfaces, to put it politely. (Really, they are very frustrating.) The Physics Toolbox Suite is also on Android. I don't know of an Android spreadsheet and graphing program like Numbers. I can provide all my data and graphs, and even encouragement and help.
