Add "wrapping it up" end for Capabilities chapter

[SeanTAllen]

Need to write a last section to chapter 5; a section that brings it all together with examples. And a big picture view of how this all fits into good pony code. Its hard to know "what does this all mean" from the individual sections that detail parts of capabilities.

[rhagenson]

We do not have a unique Capabilities chapter now -- there are two: Reference Capabilities and Object Capabilities.

I assume (without looking at the tutorial content at the time of this issue) that originally it referred to refcaps not the object capabilities. I think a wrap up section to review the concepts would be a good addition. Not sure exactly how that would look, but I suspect a bookend via an applied example would be helpful.

Refcaps are difficult to explain and (un)fortunately I have read or devised more ways to teach/learn it than there are refcaps themselves. Refcaps definitely feel like the "once you get it, you get it" concept where use leads to understanding. If we do not proceed with the wrap up example route I think the alternative is to provide a list of common use cases and match them to a set of "good" refcaps. Example, we could have this wrap up be a miniature FAQ of sorts where we present concerns like "I have data I want to share among threads/actors where each thread/actor can make changes: Ah, you want an iso of sorts" and "I have data I calculated which will never change throughout execution: Ah, you want a val of sorts" and "I have data that is private and I will want periodic updates to it: Ah, you want a ref of sorts"