Why are the metric outputs different?

Why are the metric outputs different in vOps, vFoglight and vCops. Lets start the discussion

Why is the value in vCenter ‘better’ or ‘worse’ than in vFoglight or vOPS?

It's neither better or worse; it depends on the use case. See next question.

Are there different use cases that can be solved with each?

The use case of vOPS is to primarily find "bottlenecks". Bottlenecks are usually not caused by a single blip on a graph, but rather a consistent or frequent performance degradation. The primary difference being that when someone has 100's of virtual machines to deal with, the chance of an administrator's intervention providing value is significantly higher with a bottleneck than a momentary spike. In other words, with a performance bottleneck, a configuration change will usually help the problem go away. On the other hand, if a suddenly VM spikes in CPU utilization for a few seconds and then goes away, what can one do about that?