Merge branch 'feature/5121_furyfit_fix_params' of github.com:mantidproject/mantid into feature/5121_furyfit_fix_params

Refs #5121

Conflicts:
	Code/Mantid/scripts/Inelastic/IndirectCommon.py
	Code/Mantid/scripts/Inelastic/IndirectDataAnalysis.py

Merge remote-tracking branch 'origin/feature/9205_extend_instrument_prop_of_loadmask_to_accept_idfs'

Stitch1D now looks at the min and max possible. If StartOverlap or EndOverlap are outside the min-max range, then they are forced reassigned StartOverlap=min, EndOverlap=max within the algorithm.

In the cases where we're loading IDF's without first loading in the data
(this happens for the system tests but not while using the interface under
normal circumstances), then default to the IDF with name
"INST_Definition.xml" rather than the latest IDF.

This is really just a quick fix so that we can make changes to the new
SANS2D IDF and param name that don't break the system tests.  The real
solution would be to hold off loading any IDF's at all without first
knowing which version of the instrument should be used -- this will be
done under #9361.

There were some doxygen warnings about undocumented parameters and nonexistant parameters. It was a mere case of a couple of uncapitalised Bs.

This ensures that the correct IDF and param file are loaded for old SANS2D
files as well as new SANS2D files.