Problem

flush_tier, final_flush, verbose=tiered, debug_mode.tiered_flush_error_continue, timing_stress tiered_flush_finish, drop remove_shared, and flush_tier_* / related storage stats remain in generated config and statistics after the runtime is gone.

Change

Remove those keys from dist/api_data.py and the matching stats from dist/stat_data.py (flush_tier_, local_objects, tiered_retention, tiered_work_units, dh_conn_handle_tiered). Run dist/s_all. Strip C that still parses the deleted keys (checkpoint flush_tier.enabled ENOTSUP, debug_mode.tiered_flush_error_continue, verbose map, wtperf close final_flush=true, conn_stat writers for the deleted counters).

Leave table/file tiered_storage= / tiered_object parse stubs and connection tiered_storage= parse + error-if-enabled. Do not delete WT_STORAGE_SOURCE / add_storage_source (separate ticket). Do not delete header types (tiered.h, WT_DHANDLE_TYPE_TIERED, bstorage).

test_tiered_deprecate test_flush_tier currently expects ENOTSUP. After the key is gone, unknown-config is the correct failure; update that method. Keep test_conn_tiered_storage.

Test

Ninja Debug + Python. cd dist && ./s_fast. python3 ../test/suite/run.py test_tiered_deprecate. Do not restore old functional test_tiered*.py.