That's because you're looking at the all time leagues, so the reputation change is measured from the beginning of the site. This is the same for other numbers (5,000+) except for those below 200, because, as indicated at the bottom of the widget,
users with less than 200 reputation are not tracked in the leagues
This has been fixed. The issue was in the way the "1+" group was calculated.
All the "n+" groups were exactly what they should have been, the number users with rep of at least "n", so basically "200+" includes the "500+" users.
The "1+" was instead calculated assuming the list was of non overlapping buckets, so we removed the sum of all the row counts ...
This statistic appears to be programmed to display the highest rank available between weekly, monthly, quarterly, yearly and all-time (with a bias towards the longer-term in case of ties). You can check by clicking on the profiles of the leaders on the reputation leagues.
It could be one or two things:
caching - reputation leagues are updated once a day, the /users page more often.
The 'reputation change' column in the reputation league includes reputation from so-called private reputation events: see Why is my total reputation less than my monthly reputation? on Meta Stack Exchange. For example, reputation gained on answers ...
Some explanation when something like this happens is given in this post on Mathematics Stack Exchange: Why is my total reputation less than my monthly reputation? (You might be also interested in other posts linked there, for example: What are private reputation events?)
I will quote the answer posted there:
The discrepancy is by design:
Total reputation ...
There is a problem, yet it is a bit different from the one claimed. There is no problem with the computation of the percentile.
Let us look at a site with a much smaller number of users to simplify manual checks.
The site Retro Computing has 29 pages of users (a page holds 36 names). This gives a total number of users above 1008 and below 1044. This is ...
Those numbers are the "change in rank", determined by the difference between the rank from the previous $time_frame and the rank from the $time_frame being viewed (where $time_frame is either week, month, quarter or year).
So from the image, Mike Miller is ranked #161 in the viewed year (turns out to be 2015), and saw a change (in rank) of +20 from the ...