[lePwH] ExtendableFEM.jl Simulation yields wrong results.

I have looked into it for a while so a fresh set of eyes would be helpful.
Somehow the discrete solution does not converge towards the exact solution even though the boundary conditions and weak formulation should be right.