CpuWriteGpuReadBelt for fast frame by frame memory transfers

[Wumpf]

Rewrite of #594
Introduces a convenient mechanism to allow writing directly to gpu readable memory which gpu then can copy into its internally gpu-only memory.

Goal is faster (on less memcpy), safer (alignment guarantee) and easier (no need to create buffers manually) transfers.

An unexpected ripple effect we got was that in order to have much higher memory use on Web, we need to limit the number of frames (technically: queue submissions) in flight. This makes it a lot less memory hungry as we accumlate a lot less per-frame data! Empirically, it looks a bit like wgpu handles resource cleanup a bit differently on native than on web which might just come down to the present-mode it chooses automatically. Given that the introduced limiter & device.poll makes sense regardless I haven't investigated this further.

Only used as a proof of concept in a single uniform buffer as well as point colors.

Nevertheless, some early perf numbers, need to repeat this test when we have it it more places:
(Acquired by selecting a small timeline range range and letting it loop, then opening profiler, waiting a bit, going to profiler, pause it, select a range of frames where the app was still in focus)

Nyud

DEV:
                            before   |   after (still in dev)
--------------------------------------------------------------
PointsBuilder::colors         2.280 |  2.281
PointCloudDrawData::new()     2.170 |  1.372
Overall                      12.715 | 11.762

RELEASE:
                            before   |   after (still in dev)
--------------------------------------------------------------
PointsBuilder::colors        1.784 |  2.060
PointCloudDrawData::new()    1.179 |  0.949
Overall                      9.961 |  9.748

===============================================================

colmap with lots of history

DEV:
                            before   |   after
--------------------------------------------------------------
PointsBuilder::colors         10.160 | 10.486
PointCloudDrawData::new()      6.533 |  4.209
Overall                       46.420 | 43.941

RELEASE:
                            before   |   after
--------------------------------------------------------------
PointsBuilder::colors         13.770 |  6.766
PointCloudDrawData::new()      3.682 |  2.921
Overall                       42.257 | 34.285

Btw. there is a deeper rabbit hole we ignore here: One could leverage integrated gpus (or the cache of dedicated ones) where the gpu would use the memory directly, eliminating the need to schedule a gpu memcpy, however this is behind a wgpu feature flag anyways (MAPPABLE_PRIMARY_BUFFERS).

Checklist

• I have read and agree to Contributor Guide and the Code of Conduct
• I've included a screenshot or gif (if applicable)
• I've added a line to CHANGELOG.md (if this is a big enough change to warrant it)

First half of #426

[teh-cmc]

Marvellous.

[teh-cmc]

neat

[teh-cmc]

nice save

[Wumpf]

that caused me some reeeaaal weird crash :D