[dune-pdelab] Three small requests

[Christoph Grüninger]

Dear PDELab folks,
while doing my first steps with the instationary solver from PDELab I found some minor points which could be improved.

- There is no possibility to show the assembled stability matrix. For example in dnaplfv.cc no matrix occurs at all. I suggest to introduce higher verbosity levels in the Newton solver or the OneStepMethod class which will then output the matrix, the right hand side vector and the solution vector.

- Section 5.1 of  the PDELab how-to states the availability of Runge-Kutta- and DIRK-methods which I can not find in the class documentation. Is it me, the class documentation or are they actually missing?

- The Alexander method should mention the article [Ale77] from PDELab how-to in the class documentation or is it that obvious?

Thanks for the easy way to solve instationary PDEs,
Christoph