Interesting- I guess that means Merkle's/Meyer's missed foul pop-up on the pitch before Speaker's hit is just a hair below Buckner's play in terms of WPA swings if Speakers' hit was good for a 38% .
WPA-wise you're right. Buckner's error cost the Sox 40% WE. To calculate the effect of the dropped popup in 1912, we can thank Tom Tango who helpfully provided a Win Expectancy matrix for 'close and late' situations on his site here:
WE Close and Late.
From this table I looked up the game situation with Speaker at bat (bottom of home inning, ninth or later, one out, 1st and 3rd) and the home team's WE is 46%. If the Giants get that out, it drops to 20%. So if they catch that ball, their chances increase by 26%.
But again, since the Buckner play happened in Game 6, it cost the Red Sox 20% of Championship Expectancy whereas the missed popup cost the Giants 26% - ranking it just below the Cruz play.
EDIT: For apples to apples-ness, we should probably examine both plays on the basis of what the WE would be if made or if not made. If Buckner fields the grounder (which incidentally I still expect him to do, every time I see the replay), the Red Sox are 50% likely to win that game. When he misses it, it drops to 0%. So the full effect of the misplay is -50% WPA (not simply "They were 40% to win before the play and 0% after, so -40% WPA."). The -50% WPA is comparable to the -26% WPA calculated above for the 1912 popup. Again though because of Game 6 vs. Game 7, the popup has more impact on CE - 26% vs 25%).
Doing similar "make the play/don't make the play" math on the Cruz play makes it -31% CPA.