It depends on how they did it. If they just dropped a 3.5 connector in there there is no way in hell its actually 3.5 pounds, its probably mid 4's or so.
ETA: What BillO and others still applies though, too... if someone monkeyed with the striker spring, you will want to undo that massive gayness as soon as you can.
-Mike
I'd bet its closer to 5 lbs measured at the mid point of the trigger bow.
To the OP. There are a lot of ways to measure trigger pull. Most people don't have any clue what they're talking about because they never actually compared their perception to a real trigger pull gauge. I hate to say it, but if you've never directly compared your perception with the results of a real trigger gauge, then you don't have a clue as to what the pull actually is. Until you start "calibrating" your brain against a real gauge, you will be far far off. I was, and everyone I've ever experimented with against my trigger pull gauge has been far off.
At the trigger bow midpoint, a standard glock will measure about 6 lbs. At the tip of the trigger, it will measure about a pound less.
The 3.5 lb connector (now called a 4.5# connector by glock) will get you to about 5.25lbs at the midpoint. A fluff and buff will get you to about 5 lbs.
If you drop in a 4.5 lb striker spring, heavy trigger spring, and reduced power striker safety spring, you will get that down to about 3.5 lbs.
I highly doubt all that was done. One thing the trigger spring actually pulls the trigger to the rear. So a heavier spring will make the trigger pull lighter.
I base all of this on doing probably a dozen glock trigger jobs. If you go to Wolff spring, you will see that they only charge a little more for 10 of a given glock spring than for 2 of them. So I've got a ton of springs and have played around looking for the ultimate low dollar glock trigger job.
My favorite is a 4.5# connector, 4.5# striker spring (stock is 5.5 lbs), reduced striker safety spring, and a fluff and buff. I didn't like what the extra power trigger spring did to the reset. It made it kind of non linear and did't really take much off the trigger weight.
All data given above was measured with a Lyman digital trigger gauge at the middle of the bow of the trigger.