I don't have any means to make a comparison based upon the information available other than what was already provided. To draw a definitive conclusion on the 550, we'd need the same reviewer from a trusted site to test both the G2 and the G3. That has not happened.
It may have been tested but no site that most enthusiasts consider reliable has tested both (to my knowledge), so as to provide an apples and apples comparison. One of the reasons I have no faith in Tier lists is that a model line may be supplied by different OEMs or based upon different OEM platforms.
However, here we find that the 550 and the 750 are both from the same OEM and are on the same OEM platform for the G2s as well as the G3s. It is therefore reasonable to assume that the component sourcing / quality and skill of the assembly technicians are comparable within each platform.
In the leadex series, we have:
Leadex Gold
Leadex Gold GT
Leadex Gold GTX
Leadex II Gold
Leadex Platinum
Leadex Platinum GT
Leadex Platinum GTX
Leadex Titanium
Leadex Titanium GT
Leadex Titanium GTX
EVGA doesn't make a single power supply. OEMs build PSUs for various vendors and oft two brand / models vary by little more than their decals.
For example, the Antec High Current Gamer (M) 400-620, XFX Core 450-550, XFX TS 550 Bronze, Seasonic S12II Bronze, Seasonic M12II Bronze 520-620 all vary little more than their ratings and appearance. They all were made on Seasonic's GB Bronze production line using the same componenty and built by the same technicians. Sometimes there will be slight variances by rarely of major significance as inventory control would be a nightmare. .And when it's the same vendor, this is even more true.
When ya have to worry is when ...
Corsair HX650 was made on the Seasonic G platform
Corsair HX750/850 was made on the CWT PUQ G platform
Corsair HX1050 was made on the CWT DSG platform
So in this case it would be a very sloppy inference to draw conclusions from the 750 and try and apply them to the 650 as they were made by different companies. Similarly, it would be a very sloppy inference to draw conclusions from the 750 and try and apply them to the 1050 as they were made on different platforms.
OTOH, it would be quite reasonable to draw conclusions from the 750 and try and apply them to the 850 as they were made by the same OEM on the same platform. keep in mind however that you can also see differences over time as product sourcing may change and even from two units coming sequentially off the production line