Delete settled experimental flags and dead GHG/waste fields

[briantobin-99]

cc:
Closes:

What changed? Why?

Removes five USAConfig flags whose methodology experiments are settled or that no code reads, plus their gated code:

• use_scaled_x_and_scaled_Vnorm_for_B and get_q_from_authoritative_x — the A-vs-B x-consistency experiments. Deletes their branches in derive_cornerstone_x_after_redefinition / derive_cornerstone_Vnorm_scrap_corrected / derive_cornerstone_Aq_scaled / derive_cornerstone_B_non_finetuned, the scale_cornerstone_{x,V,B} year-scalers, inflate_cornerstone_B_matrix_with_industry_pi, adjust_summary_{x,V}_dollar_year, and the authoritative-x/deflated-q helpers. derive_cornerstone_x_after_redefinition becomes the single, unconditional x pathway.
• usa_ghg_methodology and skip_scrap_adjustment_in_vnorm — read by no code at all (the scrap correction always runs; configs that set the skip flag were silent no-ops).
• eeio_waste_disaggregation — unused weights-override hook; the only setters were test fixtures duplicating the bundled default byte-for-byte. EEIOWasteDisaggConfig moves to waste_weight_config.py, where it remains the weights-descriptor type for the bundled/USEEIOR paths.

Also deletes the completed reconciling_data_years study (analysis package + its four configs; recovery point: the reconciling_data_years_final tag), full_model_scaling_for_A_and_B.yaml, compare_x_approaches.py, and analysis/q (its compare_alt_q exists to exercise the deleted q helpers).

Testing

Unit suite, black, ruff, and mypy green. None of the deleted branches are active in the v0.3 config; Snapshot integration suite (test_integration run) passes on the stack including this change.

[bl-young]

This generally looks right to me. I welcome a quick review by @WesIngwersen or @jvendries as well.