Something doesn't add up
"Surprisingly, Hutchison’s Three bests O2 for average 4G download speed in every region, which suggests network investment is paying off."
Except that in the article from Tuesday linked in the second paragraph, 3 comes in dead last. It's a bit tricky to draw any conclusions about whether investment is paying off when no-one can agree what the data actually is. You might as well use a random number generator to rank networks; the result would agree with at least one "study" and at least you'd save some time and money.