It's not clear why it should matte only because you're not representing it right. It's not saying the strikes in Afghanistan shouldn't count. It's saying that with a majority of drone strikes happen in Afghanistan, and Obama-era strikes in Afghanistan were only being counted starting in 2015, that 2243 for Trump to 1878 for Obama is not a 2-year to 8-year comparison, but rather 2 years compared to 8 years of the minority of strikes and 1.5 years of the majority.A think tank devoted to analyzing United States foreign policy "from a realist perspective" (which is to say a polite warmonger's perspective) suggests that the large increase shouldn't count because it's mostly in Afghanistan, but it's not clear why that should matter. I mean, this doesn't look like a good reason to me:
If you want to play around with the source of the numbers, you'll see they counted 0 strikes in Afghanistan prior to 2015, at some point in 2015 they started a count and found 235 strikes, and then in 2016 (the first full year), they counted 1071 strikes in Afghanistan. In 2016, they counted 58 other drone strikes, for a total of 1129 drone strikes in 2016. That is the only full year of Obama they counted. Multiply that by two to compare to 2243 for two years. They're almost exactly identical rates. The issue isn't that Afghanistan shouldn't count, but rather that it wasn't counted before. Obama only did 1878 drone strikes by them not counting 80% of the first 6.5 years of drone strikes.