feat(render): double-single fills for the face3d pipeline

Table / PolyfaceMesh / greek-text / mleader-background fills go through
the face3d pipeline; their vertices were single f32 and quantized at UTM
scale. WireModel now carries a fill_tris_low residual (filled by the
tessellate Lines path), Face3DVertex gains position_low (location 3), and
face3d.wgsl reconstructs the eye-relative position from the pair. 3DFACE
quads built from key_vertices use a zero residual (no low source).

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

src/shaders/face3d.wgsl

@@ -25,9 +25,10 @@ struct Uniforms {
 var<uniform> u: Uniforms;
 
 struct VertexIn {
-    @location(0) position:   vec3<f32>,
-    @location(1) color:      vec4<f32>,
-    @location(2) draw_depth: f32,
+    @location(0) position:     vec3<f32>,
+    @location(1) color:        vec4<f32>,
+    @location(2) draw_depth:   f32,
+    @location(3) position_low: vec3<f32>,
 };
 
 struct VertexOut {
@@ -42,7 +43,8 @@ const DRAW_ORDER_BIAS: f32 = 0.001;
 @vertex
 fn vs_main(v: VertexIn) -> VertexOut {
     var out: VertexOut;
-    out.clip_pos = u.view_rot * vec4<f32>((v.position - u.eye_high) - u.eye_low, 1.0);
+    let rel = (v.position - u.eye_high) + (v.position_low - u.eye_low);
+    out.clip_pos = u.view_rot * vec4<f32>(rel, 1.0);
     out.clip_pos.z = out.clip_pos.z - v.draw_depth * DRAW_ORDER_BIAS * out.clip_pos.w;
     out.color    = v.color;
     return out;