Appendix C: Changes to "Define Copy, move, and destroy consistently"

[eyalroz]

These changes regard "Appendix C: Discussion", section "Define Copy, move, and destroy consistently".

This section is somewhat of a jumble, mixing some fragments from other sources, and some which seem like parts of drafts of alternative versions of existing guidelines. To be honest - I would not have added it to the guidelines to begin with, and kept it on a wiki with not-yet-ready / staging material. Anyway - I would like to shape it up to better fit the general description of content in the Discussion appendix, being more tight and less meandering and fragmented.

Changes:

• Removed several notes which do not seem to add to what's already been said in existing rules, or in other text in this discussion section.
• Moved up a piece of rationale for the "rule of zero", from the discussion to guideline C.21.
• Rephrased some text to relate it to what's already been said in C.21 rather than being completely stand-alone;
• Corrected a reference to Sutter & Alexandrescu's C++ Coding Guidelines - which just said "item 53" without saying where. Plus, it's item 52.
• Added some minor emphasis to names of functions within the five, in the bulleted list of interrelations
• Removed what I believe is a false guarantee regarding classes with reference methods: "the default destructor already does the right thing"; that's not a sound promise if copying and moving has complex custom logic. Toned down the rhetoric.
• Removed the parenthesized note saying that "using a reference member is almost always wrong"; this is a strong claim that should not be made as an off-hand remark, in parentheses, in the appendix, and with no guideline even alluding as much. This fixes "Note that using a reference member is almost always wrong" ? #1707. But - see issue reference member #1809 regarding possibly guiding users to avoid reference members.

Corrected what I am certain is a mistype regarding reference methods: It is the default constructor that will certainly do the right thing.

[eyalroz]

@hsutter : It's been over a year since you self-assigned this PR. Care to review it?

[shaneasd]

The removal of the assertion that this is almost always wrong addresses #1707. My opinion is that it is therefore an improvement. However my opinion differs, it seems, from that of the editors and holds no particular sway.