forth: fix #S / UM/MOD precision bugs — Hayes 628→632/638 (99%)

Round 2 conformance fixes:
- forth-pic-step: replace float-imprecise body with same two-step
  16-bit division as # — fixes #S producing '0' instead of full
  binary string (GP6/GN1 pictured-output tests)
- UM/MOD: rewrite with two-phase 16-bit long division using explicit
  t - q*div subtraction, avoiding mod_float vs floor-division
  inconsistency at exact integer boundaries

6 failures remain (SOURCE/>IN tracking and CHAR " with custom delimiter
require deeper interpreter plumbing changes).

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

lib/forth/scoreboard.md

@@ -5,13 +5,13 @@
 | chunks available  | 638 |
 | chunks fed        | 638 |
 | total             | 638 |
-| pass              | 618 |
-| fail              | 14 |
-| error             | 6 |
-| percent           | 96% |
+| pass              | 632 |
+| fail              | 6 |
+| error             | 0 |
+| percent           | 99% |
 
 - **Source**: `gerryjackson/forth2012-test-suite` `src/core.fr`
-- **Generated**: 2026-04-25T04:57:22Z
+- **Generated**: 2026-05-05T21:30:21Z
 - **Note**: completed
 
 A "chunk" is any preprocessed segment ending at a `}T` (every Hayes test