WebGPURenderer: Improve support for output types

[sunag]

Closes #33283

Description

Improve support for different types of format and output of the RenderTarget. Previously, they incorrectly defaulted to float integer types and mappings across the pipeline layout, causing WGSL and GLSL failures during parsing or sampling.

[github-actions]

📦 Bundle size

Full ESM build, minified and gzipped.

	Before	After	Diff
WebGL	360.06 85.47	360.06 85.47	+0 B +0 B
WebGPU	633.13 175.68	633.49 175.82	+355 B +134 B
WebGPU Nodes	631.25 175.39	631.61 175.52	+355 B +133 B

🌳 Bundle size after tree-shaking

Minimal build including a renderer, camera, empty scene, and dependencies.

	Before	After	Diff
WebGL	492.3 120.06	492.3 120.06	+0 B +0 B
WebGPU	705.19 190.56	705.55 190.66	+359 B +91 B
WebGPU Nodes	654.41 177.83	654.77 177.96	+359 B +130 B

[Mugen87]

webgpu_postprocessing_dof breaks with this PR. The blur near the camera is now gone and I see these errros in the console:

I've remembered this effect since it makes use of outputStruct() to write into two render target textures at the same time (meaning the shader has two outputs).

three.js/examples/jsm/tsl/display/DepthOfFieldNode.js

Line 361 in 077dd13

const outputNode = outputStruct( nearField, farField );

Since the render target has a RedFormat, I think you should be able to fix this by changing below line although I'm not sure that's the right fix:

three.js/examples/jsm/tsl/display/DepthOfFieldNode.js

Line 371 in 077dd13

return vec4( 0 );

to

return float( 0 ); 

[sunag]

webgpu_postprocessing_dof breaks with this PR. The blur near the camera is now gone and I see these errros in the console

Thank you for checking, I was imagining that I would need to review this after removing the structure validation of the fragment, I will use this example to verify.