McLachlan set the DIrichlet scalar for gt11 etc to zero

For the Scalar boundary condition (used currently for Llaama runs since NewRad until recently would not support curvilinear outer boundaries) ML uses the parameters gt11_scalar etc to determine the Dirichlet value. I know that default values are not "sensible" values but rather "historical" values, but I question at which point in time it made sense to have the diagonal elements of the metric that we assume to become Minkowsky (since phi->0) would be zero at large distances.

Keyword: McLachlan