This worked for cs2cs / pj_transform(), but not the new API

Complements f2b3604

Assorted fixes related to +geoc flag handing

[BREAKING] Helmert: add +convention=position_vector/coordinate_frame, forbids +transpose (fixes #1091)

As identified in #1091, Helmert implementation in PROJ 5.0 and 5.1 is confusing.
It happens that by default it used the coordinate_frame convention, contrary to
the position_vector convention used traditionaly for +towgs84. The documentation
of Helmert was also wrongly specifying that the default convention was
position_vector.

This commit:
- bans the confusing +transpose parameter
- removes the concept of a default convention, since in practice both are
  equally found, and requires +convention as soon as a rotational term parameter
  is present.
  For translation only, convention is ignored and optional, as having no effect.
- fixes all the identified uses of proj=helmert in code, doc and tests

This is obviously a breaking change:
- users will have to adapt their pipeline expressions
- in particular, init files that would use helmert must be adapted

However, as designed, the break will be explicit, and not silent.

There is a regression in PROJ 5 regarding the handling of the +geoc flag,
specific to the case of +proj=longlat, and the inverse transformation,
which makes it a no-op:

echo "12   55 0 "  | src/cct  +proj=pipeline +step +proj=longlat +ellps=GRS80   +geoc +inv
 12.0000000000   55.0000000000        0.0000           inf

With this fix, we now get:

echo "12   55 0 "  | src/cct  +proj=pipeline +step +proj=longlat +ellps=GRS80   +geoc +inv
 12.0000000000   54.8189733083        0.0000           inf

The fix consists in making inv_prepare() do symetrically as fwd_finalize(), ie
skip a number of transforms when left and right units are angular,
and in inv_finalize() apply the (OUTPUT_UNITS==PJ_IO_UNITS_ANGULAR) processing
even if INPUT_UNITS == PJ_IO_UNITS_ANGULAR

Implement the Equal Earth projection (closes #1085)