Remove scripted metric aggregations from APM

[smith]

The APM service map and APM critical path feature use the Scripted metrics aggregation.

See:

• https://github.com/elastic/kibana/blob/main/x-pack/plugins/observability_solution/apm/server/routes/service_map/fetch_service_paths_from_trace_ids.ts#L58-L74
• https://github.com/elastic/kibana/blob/main/x-pack/plugins/observability_solution/apm/server/routes/traces/get_aggregated_critical_path.ts#L94-L110

We've recently been in communication with the Elasticsearch team, who has notified us:

The Elasticsearch team has come to the conclusion that we need to disable the scripted metric aggregation in our serverless offering. It's a fundamentally unscalable feature and no optimizations we do could really help with the fundamental problems with it. Similar to our reasons for disabling plugins (in serverless), we can’t continue to allow users this power if we hope to be able to manage memory in serverless (or future multi-tenant environments). Not only is the memory concern a motivating factor to disabling this - but with the move of ES|QL onto serverless, our users could start using ESQL and get better performance.

We have known issues of problems stemming from this issue in serverless, for example #178892.

We could disable these features in serverless, but our long-term plans for solving this would take significant effort and time. The service map is a GA feature. Critical path is tech preview.

We should discuss with the ES team about any other available options that could let us use ship features in serverless in a scalable way.

Make a decision about how we should handle this and update the acceptance criteria.

✅ Acceptance criteria

TBD

[elasticmachine]

Pinging @elastic/obs-ux-infra_services-team (Team:obs-ux-infra_services)

[chrisdistasio]

related issue from the past: #75366

[elasticmachine]

Pinging @elastic/obs-ux-management-team (Team:obs-ux-management)

[paulb-elastic]

@smith I think this got mis-relabelled