First of all, performance is more than running points.
Secondly, even if you look at the running results, you should compare them under the same conditions.
I have also seen an evaluation. First of all, in the test environment, the hd4000 is equipped with a cpu with a score of 19000+, and the hd5650 is equipped with a cpu with a score of 5900+. Are these two comparable? Even just looking at gpu scores will have a great impact. There are many inexplicable words in the evaluation. On the contrary, in the specific measured game, there is no comparison with the same game, and I don't know how to explain it.
In fact, it is obvious that nuclear display can't be compared with mid-range solo display, even if it was two years ago. Those evaluation articles were all written by gunmen. Just look at them and don't take it seriously.