clean up some more observables for blocks and axis

[SimonDanisch]

Improves test case:

using GLMakie

limits = Observable((-1.0, 1.0, -1.0, 1.0))

for _ in 1:5
    fig = Figure()
    for i in 1:5
        for j in 1:5
            ax = Axis(fig[i, j]; limits)
        end
    end
    empty!(fig)
end
GC.gc(true)@
@test isempty(limits.listeners) 

But still doesn't clean up if empty! is replaced by display(fig).

[MakieBot]

Compile Times benchmark

Note, that these numbers may fluctuate on the CI servers, so take them with a grain of salt. All benchmark results are based on the mean time and negative percent mean faster than the base branch. Note, that GLMakie + WGLMakie run on an emulated GPU, so the runtime benchmark is much slower. Results are from running:

using_time = @ctime using Backend
# Compile time
create_time = @ctime fig = scatter(1:4; color=1:4, colormap=:turbo, markersize=20, visible=true)
display_time = @ctime Makie.colorbuffer(display(fig))
# Runtime
create_time = @benchmark fig = scatter(1:4; color=1:4, colormap=:turbo, markersize=20, visible=true)
display_time = @benchmark Makie.colorbuffer(fig)

	using	create	display	create	display
GLMakie	3.49s (3.47, 3.51) 0.02+-	397.09ms (388.36, 401.36) 5.11+-	469.69ms (466.26, 475.99) 3.30+-	7.18ms (7.11, 7.22) 0.04+-	25.31ms (25.26, 25.38) 0.04+-
master	3.50s (3.49, 3.51) 0.01+-	363.31ms (304.72, 389.43) 38.44+-	474.47ms (463.83, 510.42) 16.27+-	7.18ms (7.11, 7.27) 0.06+-	25.29ms (25.25, 25.32) 0.02+-
evaluation	1.00x invariant, -0.01s (-1.06d, 0.08p, 0.01std)	0.91x noisy🤷‍♀️, 33.79ms (1.23d, 0.06p, 21.77std)	1.01x invariant, -4.77ms (-0.41d, 0.47p, 9.78std)	1.00x invariant, -0.0ms (-0.02d, 0.98p, 0.05std)	1.00x invariant, 0.02ms (0.51d, 0.36p, 0.03std)
CairoMakie	3.06s (3.02, 3.09) 0.02+-	318.11ms (314.65, 320.60) 2.36+-	142.58ms (140.83, 144.78) 1.58+-	7.16ms (7.07, 7.37) 0.10+-	596.95μs (593.98, 599.58) 2.35+-
master	3.07s (3.02, 3.10) 0.02+-	321.04ms (313.96, 329.04) 5.06+-	143.22ms (139.82, 151.30) 3.86+-	7.13ms (7.05, 7.23) 0.08+-	599.23μs (595.22, 603.92) 3.22+-
evaluation	1.00x invariant, -0.01s (-0.29d, 0.60p, 0.02std)	1.01x invariant, -2.93ms (-0.74d, 0.20p, 3.71std)	1.00x invariant, -0.64ms (-0.22d, 0.70p, 2.72std)	1.00x invariant, 0.03ms (0.31d, 0.57p, 0.09std)	1.00x invariant, -2.28μs (-0.81d, 0.16p, 2.79std)
WGLMakie	3.74s (3.72, 3.76) 0.02+-	318.67ms (311.96, 323.27) 3.77+-	8.59s (8.48, 8.63) 0.05+-	9.10ms (9.04, 9.17) 0.05+-	72.04ms (66.13, 80.92) 5.91+-
master	3.73s (3.70, 3.76) 0.02+-	317.61ms (313.38, 320.85) 2.95+-	8.55s (8.50, 8.59) 0.04+-	9.41ms (9.06, 11.06) 0.73+-	72.87ms (66.68, 81.51) 6.29+-
evaluation	1.00x invariant, 0.01s (0.43d, 0.44p, 0.02std)	1.00x invariant, 1.05ms (0.31d, 0.57p, 3.36std)	1.00x invariant, 0.04s (0.79d, 0.17p, 0.04std)	1.03x invariant, -0.32ms (-0.61d, 0.30p, 0.39std)	1.01x invariant, -0.83ms (-0.14d, 0.80p, 6.10std)