[Observability][Alerting] Investigate overriding default rule timeout value

[ymao1]

With this PR, the alerting framework is now allowing individual rule types to override the default alerting task timeout of 5m. (Note that this default value is now also configurable via the kibana.yml config. The framework will be gathering telemetry on rule execution durations before deciding whether to extend the default alerting timeout value but in the meantime, individual rule types should consider whether they want to provide a longer override value.

Things to consider:

• does this rule type have a history of poor performance (prior SDHs)
• is there a use case for this rule type to run against frozen indices

Note that we do have an issue open to potentially allow individual rules to override the timeout but for now we are just enabling at the rule type level.

[ymao1]

When implementing this, it might be nice to have these values by configurable in the kibana.yml or advanced settings so they can be changed between releases if necessary

[elasticmachine]

Pinging @elastic/kibana-alerting-services (Team:Alerting Services)

[elasticmachine]

Pinging @elastic/infra-monitoring-ui (Team:Infra Monitoring UI)

[ymao1]

Closing as duplicate of #123053