fix(nextclade): align dataset metadata and qc config with schema

[ivan-aksamentov]

Fix two categories of schema defects in the mumps Nextclade dataset configs.

Metadata flags: deprecated, experimental, and enabled are at the top level where they are silently ignored. The experimental flag belongs under attributes. The enabled field is not a valid dataset-level property.

QC scoreWeight (genome only): scoreWeight on qc.missingData and qc.mixedSites is not defined in the schema and is silently ignored. It is valid on snpClusters, frameShifts, and stopCodons (those are left unchanged).

Field	Action	Datasets
deprecated: false	Remove (default is false)	genome, sh
experimental: true	Move to attributes.experimental	genome, sh
enabled: true	Remove (not in schema)	genome, sh
qc.missingData.scoreWeight	Remove (not in schema)	genome
qc.mixedSites.scoreWeight	Remove (not in schema)	genome

• Move experimental into attributes in both datasets
• Remove deprecated and enabled from both datasets
• Remove invalid scoreWeight from genome QC config

⚠️ The existing datasets in nextclade_data are already patched in nextstrain/nextclade_data#437, so no immediate resubmission is needed. But without this upstream fix, the issues will reappear on the next dataset submission from this workflow.

ℹ️ See pathogen.json schema for the PathogenJson, PathogenAttributes, QcRulesConfigMissingData, and QcRulesConfigMixedSites definitions.

👀 Note: #40 adds QC config to the SH dataset and uses the same invalid scoreWeight on missingData and mixedSites. Those fields should be omitted when #40 is updated.

1. Related to: fix: update mumps dataset schema fields nextclade_data#437