There is no good answer to this question as the final choice of an MT engine might come down to many possible factor. For example, some MT engines produce better results for a specific language pair or if equally capable, results might be affected by the topic. Some companies publish some analysis of MT output and rank engines for some language pairs. One such company is inten.to, and they post some results on a regular basis. The latest study is available here.
Studies like this one might give some insights into what engine is better for some language pairs, but when dealing with a large project, users should perform some testing with their translation teams and collect some feedback. The perception from translators might vary from the automated results generated for the studies. Testing a couple of thousand words with selected engines should be sufficient in most cases.
