Fixes permitting digitization to run in the sqlite-based detector description configuration.

There is a simple bugfix in SCT digitization, and more significant changes in Muon code in view of getting to functional digitization using the new detector description coded:

• The problem was: Since the Muon geometry is rebuilt in MuonDetectorCondAlg, a new geometry tree is created, and components from the old tree are copied to the new tree. But in the new sqlite configuration, the tree is prebuilt. In that case old and new trees are the same and copying-to-self takes and infinite amount of time (size of self increases indefinitely). The solution is: This is fixed by checking for copy-to-self. Other changes in muon code affecting only the sqlite configuration were also required for the muon code migration.

• The problem was: In the MM_Digitization tool, readout geometry information was fetched from the AGDD service. This service does not run in the new DD. The solution is: Instead, the information is taken from the RDBAccessSvc, which ultimately come from SQLite inputs.

The latter change is implemented in sort of a ham-handed way for the moment; this in the spirit of incremental MR's that are not very big. Very shortly after (as explained in the code comment) the missing information will be filled at initialization time rather than in the digitization tool, in a new MR. And, this currently does not have any impact outside of the sqlite configuration.