True, and I don't know what the rationale behind their specific motivation is, but if I were to make a website, I'd do the same. That's not to say that they shouldn't be more transparent about their methodology. The truth of the matter is that there's very little known about the people that actually die in these strikes.

There is a fair amount of confusion on whether the deaths reported to be terrorists are actually terrorists. Some say they are not especially since the dead are never charged. There are also reports that the Pakistani military straight-up calls civilians terrorists[0] but these are heavily contested by the military[1] itself.

If you had a dataset which was heavily reported to be mislabelled, what would you do?

Personally, I'd do the same but perhaps add a better explanation.

[0]: https://www.bbc.com/news/world-asia-48139648

[1]: https://www.dawn.com/news/1486313