docs: one changelog line for the viewer's grid-scan mode, not two

The mode arriving and the mode becoming settable are one feature to a reader, and
they landed a few hours apart. Folded together, as the rest of the section already is.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01EFEJG6WBQv8th4UJFNe53N

docs/CHANGELOG.md

@@ -8,8 +8,7 @@
 * Rotation MX analysis is not available in jfjoch_broker and is processed with rugnux; every other method carries its own settings on its own endpoint, `/config/grid_scan_analysis` and `/config/calibration_settings`, and `PowderCalibration` computes the azimuthal profile on the CPU with a sector count of its own.
 * A grid scan finds the crystals in the finished map and reports them - grid position, nearest image, measured size and orientation, score, ice score and resolution - through the CBOR end message (`grid_crystals`), the HDF5 master (`/entry/MX/crystals`) and `GET /result/scan`; `/config/grid_scan_analysis` governs what counts as one - the protein-score threshold, the grow threshold, the minimum cells per crystal, the decisive single-cell score, the crystal cap, and whether each cell is indexed, which it is by default.
 * jfjoch_viewer draws a grid scan as one map - protein in orange, ice in cyan, mixed where both are present, over the shape of the loop - with a frame around each crystal found, at its own orientation.
-* jfjoch_viewer runs the grid-scan analysis itself: the settings panel offers Grid beside MX, AzInt and Calib, and "Analyze dataset" scores the whole raster and draws the crystals it found on the map.
-* The jfjoch_viewer Grid page carries the grid-scan analysis settings, the unit cell, spot-finding and indexing sections are shown for it as well as for MX, and the raster geometry - step, points along the fast axis, vertical and snake - can be stated in the Goniometer section for a file that records none, marked as edited where it is not the file's.
+* jfjoch_viewer runs the grid-scan analysis itself: the settings panel offers Grid beside MX, AzInt and Calib, "Analyze dataset" scores the whole raster and draws the crystals it found on the map, and the Grid page carries the analysis settings alongside the unit cell, spot-finding and indexing sections; the raster geometry - step, points along the fast axis, vertical and snake - can be stated there for a file that records none, marked as edited where it is not the file's.
 * `rugnux --mode raster` runs the same analysis over a stored grid scan and writes `<prefix>_raster_report.txt` and `<prefix>_raster.json`; `--beam-size` states the beam the reported sizes still contain, and where neither it nor the file says, the grid step is reported in its place rather than zero.
 * The per-image ice ring measure is renamed `ice_ring_ratio` (`/entry/MX/iceRingRatio`, plot type `ice_ring_ratio`); the retired `ice_ring_score` spelling is still read from older files and streams, and still accepted by the API.
 * `rugnux --model` reports CC(model, data) by resolution shell on the same shells as CC1/2, with the reflection count and a significance for each, and fits the model's scale, anisotropic B and bulk solvent on the working reflections only, so the R-free is measured against a model no free reflection helped scale.