Momentmodels, bindable numerical fluxes, etc

[tobiasleibner]

This is #157, transferred to a new branch to see if that fixes CI.

[codecov]

Codecov Report

Merging #170 into master will decrease coverage by 1.43%.
The diff coverage is 80.21%.

Impacted Files	Coverage Δ
dune/gdt/local/finite-elements/lagrange.hh	76.36% <ø> (-1.28%)	⬇️
...t/test/momentmodels/entropyflux_implementations.hh	65.97% <ø> (ø)
dune/gdt/type_traits.hh	0% <ø> (ø)	⬆️
dune/gdt/local/integrands/product.hh	63.8% <ø> (+23.84%)	⬆️
dune/gdt/local/bilinear-forms/integrals.hh	85.67% <ø> (+15.75%)	⬆️
dune/gdt/spaces/interface.hh	19.46% <0%> (-2.2%)	⬇️
dune/gdt/local/functionals/integrals.hh	54.44% <0%> (-21.54%)	⬇️
dune/gdt/local/finite-elements/interfaces.hh	64.51% <0%> (+0.77%)	⬆️
dune/gdt/local/numerical-fluxes/vijayasundaram.hh	0% <0%> (-100%)	⬇️
...local/assembler/operator-fd-jacobian-assemblers.hh	0% <0%> (-62.24%)	⬇️
... and 264 more

[tobiasleibner]

The issues from the last merge should be fixed, so this is again ready to be merged.

[ftalbrecht]

I left a bunch of comments. In addition, we should talk about the discrete/nondiscrete operator :/

[tobiasleibner]

@ftalbrecht, I addressed your comments and there are some new open issues now for the parts we decided to do after the merge. Please have a look at the new commits and feel free to merge this PR if you are okay with these changes.

[ftalbrecht]

A few virtuals seem to be still there, apart from that I am happy!

[ftalbrecht]

The virtual seems to be still there

[tobiasleibner]

Yep, my search and replace did not take into account functions where virtual and override are on different lines. I will remove the remaining virtuals manually whenever I see them.