You are correct. The relevant section is this (emphasis theirs):
All of that is for an average income family. Low-income children benefit from starting earlier, and high-income children from starting later. (It’s likely not just actual income that matters, but socioeconomic factors — but income is easy to measure objectively so researchers measure it.) The most deprived children can actually benefit from starting as 1-year-olds.
You are correct. The relevant section is this (emphasis theirs):
I correct the TL;DR.