Preserve direct light specular reflections for smooth materials

[SeanCurtis-TRI]

This tweaks the mathematical model so that perfect reflections of our direct lighting don't dwindle away to nothing (counter to how they would behave in the real world).

---

This change is 

[SeanCurtis-TRI]

+@jwnimmer-tri for feature review, although feel free to dispatch this to someone else.

Note: There's no specific unit test. The tutorial uses it, so that is, vaguely, unit-testy. Let me know if you'd like an explicit unit test. I'm pretty sure I can come up with one.

Reviewable status: LGTM missing from assignee jwnimmer-tri(platform), needs at least two assigned reviewers, missing label for release notes (waiting on @SeanCurtis-TRI)

[jwnimmer-tri]

I don't think we need to add a unit test in this PR. Let's circle back after the tutorial lands and revisit if we want to backfill one.

Reviewed 3 of 3 files at r1, all commit messages.
Reviewable status: 2 unresolved discussions, needs at least two assigned reviewers, missing label for release notes (waiting on @SeanCurtis-TRI)

---

tools/workspace/vtk_internal/patches/preserve_direct_light_specular_reflections.patch line 13 at r1 (raw file):

completely).

In this patch, we're simply mapping the roughness domain [0, 1] to [0.05, 1.0]

BTW We should probably document our reasoning / heuristic / taste for choosing 0.05 vs 0.02 vs 0.08 etc. Is there anything principled? Or are we just looking at 1 (10? 100?) sample render and using our aesthetic judgement? Or are we matching Meshcat's floor?

---

tools/workspace/vtk_internal/patches/preserve_direct_light_specular_reflections.patch line 50 at r1 (raw file):

+  // remapping input roughness [0, 1] to the range [0.05, 1.0], guaranteeing
+  // that lights create a specular highlight on the smoothest surfaces.
+  return 0.95 * roughness + 0.05;

BTW Do we want 0.95f and 0.05f to keep the compute in float-precision? Maybe it doesn't matter.