Document the parameter list each destination type takes

[MahathirMohammadShuvo]

What kind of change does this PR introduce?

Documentation. This is the note you said yes to in #1783.

docs/destinations.md showed three worked calls but never stated the parameter lists, and
six of the eight destination types appear nowhere on the page — so for those there was
nothing to infer an arity from at all. The note adds the eight types from ISO 32000-1
Table 151 with their positional parameters, which of them accept null, and what the
zero-argument call produces.

It also names the one conversion the method performs, which is the part I think earns its
place: an XYZ top is flipped against the page height, while FitH, FitBH and FitR's
vertical coordinates are written through unchanged. The page's own two examples already
disagree about the origin —

doc.addNamedDestination('LINK', 'FitH', 100);        // -> [page /FitH 100]
doc.addNamedDestination('LINK', 'XYZ', 36, 36, 50);  // -> [page /XYZ 36 756 50]

— and nothing said so, so a reader who carries the XYZ example's convention into a FitH
call lands at the wrong end of the page.

Every arity and the null rule were checked against the pdfjs-dist in this repo's dev
dependencies, and FitR's parameter order against the way its viewer consumes them
(x = destArray[2], y = destArray[3], width = destArray[4] - x).

One thing worth flagging rather than leaving you to find: the table will render as literal
pipes, because the docs are built with markdown@0.5.0 and the table dialect is not
enabled. docs/getting_started.md and docs/outline.md already ship that way through both
generators, so I matched the neighbours rather than making this page the odd one out — but
say the word and I'll convert it to a bullet list.

I'll close #1783 in favour of this.

Checklist:

• Unit Tests — n/a, documentation only
• Documentation
• Update CHANGELOG.md — n/a, the Unreleased section carries behaviour changes
• Ready to be merged

[blikblum]

Many thanks