I don't think it is a good idea, because the math doesn't work out:
FFG SW RPG(s)
In the FFG Star Wars series of RPGs, the success chance is scaled with 3 elements: Type of die (yellow, green, blue vs. red, violet, black), the number of dice (that is for each die) and then the available upgrades to either side. As a result, you get beautiful bell curves, averaging out in the middle field around some point. While you can pretty easily determine if a success is possible at all with a certain pool, it is more problematic to determine the chances for any of the possible events.
A shot with 1 yellow, 2 green vs. 2 violet is a pretty good chance to get some success but it could fail as horribly as it could win glamorously. Calculating the exact chances for all the possible outcomes is hard, but luckily doable with computer and anydice. In this case of stat 3 + skill 1 (pretty much what a stater char can manage) we have a chance of 65.08% to get at least 1 success. The upgrade of a green gives us 70.6%. No chance for a fumble, 8.33 % / 15.94% to get one (or two) triumph(s).
Pathfinder, on the other hand, scales very different: it scales linear because there is only one random number plus a static modifier. 1d20+6 for example. This is a pretty linear scaling: as long as your modifier is large enough, you can't fail but for a 1 (which technically only is a fumble on trying to use magic wands), if it is not high enough, you can't make it but for a natural 20 (again, RAW only on attack).
Shooting with a +5 at an AC 15 monster (pretty much not hardcore built lvl 1 picture) is a 50% chance, the chance for a confirmed crit is 2.5%. Not more, not less. Getting a fumble on a nat 1 is actually a Houserule that has been imported from D&D 2nd!
Shifting the AC down to 14 or adding a +1 modifier shifts the chance to hit up by 5% to 55%. At the same time it makes a crit not that much more likely (5% to get a chance, times .55 for a confirm = 2.75%).
I don't think the two systems are to be made compatible without breaking either of them.