I did something similar, because I wanted to see if Jason's 3A was really as much of a problem as a lot of people say it is. So I looked at the stats from 23 different skaters for this season and last (though it's not the exact same skaters for those 2 seasons), and counted the 3As either as +GOE, UR, pop, fall or completed (meaning +GOE and -GOE but not caused by one of the previous three mistakes).
(I also had the same problem with combinations... I ended up counting 3A combos were the mistake happened on the second jump as completed, but not as +GOE)
I counted only international events though, since for a lot of skaters getting the national scores is, well, a pain in the ***.
Sorry!
And, at least from that group: Jason's succession rate last season was the third lowest (behind Misha Ge and Ross Miner, who usually has a pretty good 3A) with 42,8% getting +GOE and 64,3% being completed. The average is 62,3% with +GOE and 83,6% completed for all the guys. This season, he had +GOE on 50% of his 3As, while the average is 58,8% (when it comes to the +GOE, he's had a better succession rate this season than Artur Gachinski, again a guy with a usually very good 3A, and the allegedly consistent Maxim Kovtun). So first of all, there's an upwards trend, and second the difference is not as big as I think some people make it sound.
Also he tends to make most of his mistakes at the beginning of the season while a lot of the other guys are all over the place or tend to be even better in the first half. So I'd rather expect his stats to go up compared to last season again after worlds - and IMO the fact that he's still improving is the most important info from this
It's also pretty much what you already said before: he URes, but falls and pops less then others. His UR rates are the highest out of all the skaters - last season it were 28,6%, this season so far it's 25%. It's more noticeable this season though because there is only one other skater to get any UR call on the 3A (Misha Ge, and it was just once). Last season there were a lot more (Aaron, Brezina, Kozuka, Miner, Takahashi and Yan for example - all of which are very good 3A jumpers overall).
If you look at falls and pops combined, Jason was at 7,1% last season and had a lower mistake rate than Abbott (21,3%), Brezina (15,4%), Dornbush (8,3%), Fernandez (10%), Ge (25%), Hanyu (10,5%), Kozuka (8,3%), Menshov (12,5%), Miner (33,3%), Mura (8,3%), Rippon (22,2%), Voronov (11,1%), Ten (11,1%), Yan (16,6%) and Chan (11,1%). Only 7 guys were better than Jason in the no fall and no pop category. This season looks different since he had 2 falls on the 3A in the earlier comps. But since the overall succession went down, he's more middle of the pack now in that regard.
Base line: His 3A is giving him more trouble then it does for the "average" skater. But IMO, it's not as bad as it's made out to be and he has the bad luck of competing under a system were URs are worse than falls. He's also still improving, so that's good news
PS: it's off topic but my fan heart needs to say this: Oddady kicked everyones behind with a +GOE rate of 100% last season! And with the second highest average GOE of 1,93 no less. Now I miss him again...