Why the drastic discrepancies re instances?

I was putting together a Facebook post today for Isaac Watts' 350th birthday, and I used the list of his hymns on his person page to get a list of his 8 most-published hymns (the criterion was actually “1000 or more instances”). I opened the hymn authority pages and was immediately confronted on the number one hymn by a drastic between the number of instances in the list on the person authority page and the number of instances given on the hymn authority page: The person-authority list I had used to get my eight winners said there were 1833 instances of “Alas! and did my Saviour bleed?”, but the hymn authority said there were 2309 instances of it! A difference of almost 500 instances! I'm reconciled to the large number of hymns that show 2 instances in the lists but only 1 in reality, but 450+!? What causes this sort of thing??? It's not just that one hymn. The other seven hymns on the list show…

When I survey the wondrous cross: 1654 v. 2001

Joy to the world: 1559 v. 1863

Am I a soldier of the cross?: 1541 v. 1946

Jesus shall reign where'er the sun: 1482 v. 1801

Come, w[/y]e that love the Lord: 1420 v. 1791

There is a land of pure delight: 1022 v. 1403

O[ur] God, our help in ages past: 1014 v. 1241

and to take a less frequently reprinted hymn of his, 

Behold the glories of the Lamb: 193 v. 276

I just find this hard to fathom. And the discrepancy isn't a fixed percentage; “There is a land of pure delight”, no. 7 in the list, has the second largest discrepancy (381 to “Alas!”'s 476). It's quite likely that if I went by the hymn authorities instead of the person authority I might find others that crept upward into the top 8, displacing who knows which.