That's exactly what this graph (which I posted on the previous page) shows. 18 different analyses by different, independent groups, each using their own statistical techniques (to decide how much area each site represents, etc.), some with homogenisation, some without. Do you see any massive differences worthy of crying conspiracy over?
