I'm not an expert on "big data manipulation", but I know my way around Excel pretty well. So I took on Jerry and David's challenge, and here are the results.

There are just over 2.2 million QSOs in the database, but that includes non-US stations/grids, and QSOs that don't include grids. Filtering to just QSOs with US stations with grids included and grids within the 436 that count for the 070 award takes the total down to about 598 thousand. I then tallied (not manually!) how many QSOs were with each of those 436 grids. Note that this counts QSOs, not unique callsigns (so there are many duplicates in the count). I also did not do any real quality checking other than that they were valid 4-character grids that matched one of the 436. So if someone had logged me at my home QTH (which is FN20), but put EN20 or DN20 as the grid, it was counted as that errant grid.

- There are eleven grids that have no QSOs at all: 

DL88

DM85

DN01

DN02

DN25

DN34

DN52

DN57

DN65

DN75

EN28

- There are thirty other grids with less than ten QSOs each:

DN11	1
DN12	1
DN67	1
DN92	1
DM28	2
DM39	2
DN21	2
DM47	3
DM48	3
DN04	3
DN85	3
FN65	3
DM17	4
DM38	4
DM53	4
DM74	4
DM86	4
DM87	4
DN51	4
DN87	4
DM19	5
DM90	5
DN56	5
DM96	7
DN10	7
DN93	8
DN95	8
EN38	8
DN53	9
DN68	9

- There are four grids with over 10,000 QSOs each. 

FN42	10027
FM19	10262
FM18	12395
FN20	20760

I'll be posting the results spreadsheet to the Yahoo page if anyone wants to check on other grids.

Steve

W3HF