Or may be too lazy to express themselves properly, and never improve cause karma whores spoil them with answers? :)

Anyway I learned new things again, I wasn't aware that the "NP-heuristics" are able to guaranty such good results like Christofides does.

Thats impressive, especially cause NP-problems are transformable into each other.

(Though it might not help much, if one needs to know for 100% if two graphs are isomorphic or not.)