I finally got around to doing stuff with that data. In the below graphs the blue line is the actual attendance markings, while the red line is the "moving average", it starts after the 9th home game in order to avoid extremely small sample sizes.
For those wondering, the top 10 attended games:
Attnd. | Opp. | Month |
34805 | BAL | August |
36048 | BOS | April |
34145 | BOS | June |
36048 | BOS | July |
34441 | CHC | June |
36048 | CHW | May |
36048 | DET | August |
36048 | MIN | September |
36048 | MIN | September |
36048 | SEA | April |
And the bottom 10:
Attnd. | Opp. | Month |
8269 | TOR | April |
8989 | TOR | April |
9540 | TOR | April |
10511 | TEX | May |
10927 | TEX | May |
11898 | SEA | April |
12039 | LAA | May |
12106 | SEA | April |
12146 | BAL | April |
12174 | TEX | May |
Just from these two tables it seems the attendance tide shifted sometime in May.