Firstly, the curves were originally made for formula feeding babies, so whilst they might have been changed from their original data points to include bf babies, that was the original intention for them. As you're right, breastfed babies can't be overfed.
Secondly, I was told that in-utero babies grow to the placenta's ability to feed them. Once born, their genetic makeup takes over as to whether they will become a "big baby" or a "small baby". Two petite parents can have a whopping size baby, but ultimately this baby may be destined to be short and "smaller than average" as that is what it is genetically programmed to be. These babies fall through the height/weight charts (which was my baby!). Your baby might have been destined to be bigger, and rather than being born big is making up for it now, so to speak. Someone has to be on the 100 and someone on the 1. Both are healthy, but statistically they are rarity, and average size is more common.
Thirdly, growth where it is matched on both length and weight is usually OK, it's when there is a mis-match they might have concerns (weight climbing through percentiles, but height following the curve).
And lastly, breastfeeding studies had women express milk before and after each feed for 24hrs, as well as weighing baby pre and post feed. Based on this, measurements were made of the amount of milk a baby drinks in 24hrs and also the fats they receive in 24hrs. For many hundreds of data points, the amount drank was massively different - but the fats they consumed was almost exactly the same. Babies that feed more frequently get a less rich breastmilk, and babies that drank less got more fats in each feed.
So if baby seems happy, relax and keep feeding - it will all be a distant memory when they are starting school and noone could pick who was on what centile when.