Dont adapt Flat direction advection scheme plus better summary for FluxFormAdvection

[glwagner]

Before this PR, a simple two-dimensional turbulence simulation emits the following display:

julia> model = NonhydrostaticModel(; grid, advection=WENO(order=5), tracers=:c)
[ Info: Using the advection scheme Upwind Biased reconstruction order 1 in the x-direction because size(grid, 3) = 1
NonhydrostaticModel{CPU, RectilinearGrid}(time = 0 seconds, iteration = 0)
├── grid: 128×128×1 RectilinearGrid{Float64, Periodic, Periodic, Flat} on CPU with 3×3×0 halo
├── timestepper: RungeKutta3TimeStepper
├── advection scheme: FluxFormAdvection{3, Float64, WENO{3, Float64, Nothing, Nothing, Nothing, Nothing, WENO{2, Float64, Nothing, Nothing, Nothing, Nothing, UpwindBiased{1, Float64, Nothing, Nothing, Nothing, Nothing, Centered{1, Float64, Nothing, Nothing, Nothing, Nothing}}, Centered{1, Float64, Nothing, Nothing, Nothing, Nothing}}, Centered{2, Float64, Nothing, Nothing, Nothing, Centered{1, Float64, Nothing, Nothing, Nothing, Nothing}}}, WENO{3, Float64, Nothing, Nothing, Nothing, Nothing, WENO{2, Float64, Nothing, Nothing, Nothing, Nothing, UpwindBiased{1, Float64, Nothing, Nothing, Nothing, Nothing, Centered{1, Float64, Nothing, Nothing, Nothing, Nothing}}, Centered{1, Float64, Nothing, Nothing, Nothing, Nothing}}, Centered{2, Float64, Nothing, Nothing, Nothing, Centered{1, Float64, Nothing, Nothing, Nothing, Nothing}}}, UpwindBiased{1, Float64, Nothing, Nothing, Nothing, Nothing, Centered{1, Float64, Nothing, Nothing, Nothing, Nothing}}}
├── tracers: c
├── closure: Nothing
├── buoyancy: Nothing
└── coriolis: Nothing

After this PR we see

julia> model = NonhydrostaticModel(; grid, advection=WENO(order=5), tracers=:c)
NonhydrostaticModel{CPU, RectilinearGrid}(time = 0 seconds, iteration = 0)
├── grid: 128×128×1 RectilinearGrid{Float64, Periodic, Periodic, Flat} on CPU with 3×3×0 halo
├── timestepper: RungeKutta3TimeStepper
├── advection scheme: WENO(order=5)
├── tracers: c
├── closure: Nothing
├── buoyancy: Nothing
└── coriolis: Nothing

if the grid is Flat in the vertical. Also if we do adapt the advection order, it cleans this up:

julia> model = NonhydrostaticModel(; grid, advection=WENO(order=5), tracers=:c)
[ Info: Using the advection scheme UpwindBiased(order=1) in the x-direction because size(grid, 3) = 1
NonhydrostaticModel{CPU, RectilinearGrid}(time = 0 seconds, iteration = 0)
├── grid: 128×128×1 RectilinearGrid{Float64, Periodic, Periodic, Bounded} on CPU with 3×3×1 halo
├── timestepper: RungeKutta3TimeStepper
├── advection scheme: FluxFormAdvection(x=WENO(order=5), y=WENO(order=5), z=UpwindBiased(order=1))
├── tracers: c
├── closure: Nothing
├── buoyancy: Nothing
└── coriolis: Nothing

[navidcy]

Yes!!