pointsInstance.

For now also dump topology related sets to pointsInstance since otherwise
paraFoam cannot see them (only looks at first 'sets' directory when searching
back)

This comes up when displaying sets on moving meshes. The topology does not
change so the set should stay the same.

- Earlier calculation of dQ was misleading - gave a measure of temperature
  change and not heat release - also mixed molar/mass units so value was
  somewhat random
- Updated to calculate enthalpy/sec [m2/s3] and improved/more descriptive
  comments

- Left over from earlier commit 2900792f

Removed #include file

- added directMapped*Patch constructors with uniform offset
- changed splitMeshRegions to use it
- adapted chtMultiRegionFoam tutorials

- info now written as strings
- info now collected on master processor and written to top level
  postProcessing folder as postProcessing/lagrangian/<cloudName>/<time>/...

The refValue was being converted by mixedEnthalpy (even though it is not
used) so a sensible value is necessary to get it to go through the conversion.
So it now just gets the value

The boundary conditions were still accessing e.g. the temperature even
when restarting. Now they
will reread all their data if there is a 'value' field present.

An empty list (m_) has no pointer so cannot be referenced. Instead we
now return a null pointer (which is no problem for UList)

- maxStoredParcels was not being initialised

- Added call to Particle<Type>::readfields(CloudType) in derived types