Google says 47.9%, but obviously some teams are going to be better at it than others, and I'm just guessing from the general state of the Giants they're probably not one of those teams.
When losing by 8 points like the Giants were after their TD, your odds of completing the 2-pt conversion only have to be higher than 39.2% to give you a better chance of winning than kicking the XP.
It was the correct call and it should be common sense for teams to go for 2 when they are losing by 8 in the 4th quarter. The analytics behind this are quite simple.
Let's say that a team gets 40% of their 2-pt conversion. The team is losing by 2 TDs in the 4th quarter. If we assume that the team succeeds into getting two TDs, it should absolutely go for 2 after the first TD.
Doing so, they have 40% chances of trailing by 6 after the first TD. In that scenario, they kick the XP after the second touchdown and win by 1.
However, they'll also be trailing by 8 60% of the times since they missed their conversion. Obviously that is not good, but they still get a chance to tie the game by going for 2 after the second TD. In fact, their chances of having a tie game after their two TDs are 60% * 40% = 24% and their chances of losing are 60% * 60% = 36%.
In short, assuming that the team succeeds in getting the two TDs (and obviously that the other team does not get any additional points), if they go for 2 after the first TD, they have 40% chance of winning, 24% chance of going to overtime and 36% chance of losing the game. Those are better odds of winning the game than kicking the XP and sending the game to overtime.
And that is even if you are bad at getting 2-pt conversion (40% chance of doing it). With odds of 48% of getting the 2-pt, the odds change to 48% of winning the game, 25% of going to overtime and 27% of losing.