Theoretically speaking, T2 can never be greater than 2*T1. (ref: https://dtrx.de/od/mmm/mmm_20180117.pdf). But why T2>2*T1 components often appear in NMR T1-T2 correlation maps?
A very simple way to understand this is T1 involves magnetization in the Z domain/dimension, whereas T2 involves magnetization in the X-Y domain / two dimensions. Further, since T2 experiments begin by transferring magnetization from the Z axis to the X-Y axis, T2 amplitudes cannot be independent of T1 amplitudes.
It is because in 3D systems, the realm of X, Y, and Z are equally "spaced." Thus in the T2 domain, assuming the spacing of the X and the Y realm is equal, of course data collected for both X and Y would be double that of X and also the double that of Z. T1 data in Z is only in one dimension, and T2 data is always in two dimensions.