I took a look at the data again and it just appears that the bandpass filter may not be correct for the data. The filtered data has a very distinct peak at the start that if you want to ignore, you will have to program the Code to ignore the first few points in the Max.
What is this data, and might that help us figure out why the filter doesn't seem right? I also notice that when you added the cursors you did not associate them with a specific plot. If you want to click on the lock on the cursor legend to associaate that cursor with a plot.