Specifically the last 4 digits in the 7 and 8 digit numbers repeat far more often than expected by chance. For example, the extent numbers end in "0156" 59 times (as of today). "0938" comes up 54 times. Etc. These all occur much, much higher than chance (which is less than once in a data set this size), and suggest there might be a small systematic error in the data.
Weird. I've asked JAXA about it.