So it seems like this is a complicated measurement of latency. The CDNs should win where they have a nearby endpoint. A comparison of the additional latency (d

So it seems like this is a complicated measurement of latency. The CDNs should win where they have a nearby endpoint. A comparison of the additional latency (due to slow webservers etc) seems like a more relevant thing to measure if we'recomparing the services themselves. Eg something like time to first byte MINUS network latency. Otherwise you already had your answer with the ping times..