Today is NPD day, a day that is usually good for some high drama on social networks and message boards. As I look over the data, however, I feel like I'm getting about half of the picture. There are so many inconsistencies that make for good press releases but don't actually mean anything in the real world.
The biggest problem with the NPDs is that they don't count digital or bundle sales. Sony trumpeted that it had sold the most copies of Madden, but did it really? With EA Access discounts, and the game already being loaded digitally due to the free playtime, it's possible that Madden sold quite a few copies on the Xbox One digitally. There also was a Madden bundle, which was in short supply, but those copies weren't counted either. So while Sony sold more hard copies of the game through NPD tracked retail outlets, I don't think we can decisively say that it outsold the XB1 version.
Likewise, there were more Xbox One hard copies sold of Destiny than PS4 copies. What this leaves out, however, is every copy that came in the Destiny Bundle which has been sitting at the top of various online retailer charts for the last few months. There was also a large push on the PSN to pre-order and pre-load digitally. Without those numbers, can we say there are more Xbox One Destiny copies in the US than PS4? There just isn't enough data to make that statement conclusively.
There are certain things the NPDs are decent for. Seeing something like 3DS Smash Bros at Number 4, with only two days worth of sales on the charts, versus nine other games that were multiplatform, gives us a good indicator that the title sold well. Nintendo, in a press release of their own, confirmed that by saying that 705,000 copies were sold in those two days. So the NPD, in this case, did give us some solid evidence that turned out being right.
The NPD is also slightly useful for games that aren't bundled and aren't likely to be massive digital sellers. A game like Shadow of Mordor, while I own it digitally, is the type of game that most people buy the physical version because the lack of multiplayer makes it more likely you will be trading it in after beating it to help fund further purchases. If you go with that mindset, seeing PS4 having sold more than Xbox One is *probably* right but, again, how can we be totally sure?
The utter lack of hard data in the NPDs, and the complete exclusion of other data, makes the report pretty useless. None the less, it is what people cite as "facts" when getting into system related arguments. You know what probably made the most money last month? League of Legends, DoTA 2, World of Warcraft, and Hearthstone. When was the last time you saw those games show up on a NPD? So while Xbox warriors will trumpet the "higher sales" of Destiny on their platform, and PS4 warriors will trumpet the higher sales of Madden on theirs... their "facts" could be completely false.
Depends on if there are press releases for the specific titles you mentioned. For example, both Activision and Sony have press releases touting PS4 as #1 platform for Destiny sales. They probably have the full picture since they can directly track their digital and bundle sales as well as NPD. If there was a press release from MS claiming Destiny sold most on their system then NPD would be accurate. For everything else though you're completely right, it doesn't give us the full picture. Another thing is that people are so focused on NPD that sales in other territories become less important to the discussion, but certainly not to the bottom line of the hardware companies and publishers. The most concrete thing NPS is good for is North America hardware sales.
Ever since digital distribution of software became common place, NPDs are no longer a reliable metric for software sales. I would still trust them as a reasonable metric for hardware sales, though.