chore(eks): graduate experimentalInPlaceUpdates once validated against a live EKS cluster

[devantler]

🤖 Generated by the Daily AI Engineer

Evidence

#6195 ships the EKS Updater capability (managed node-group scaling in-place) behind the default-off typed config field spec.cluster.eks.experimentalInPlaceUpdates, because the path could not be validated against a real EKS cluster before merge. The portfolio flag-lifecycle rule (feature-flag-first, monorepo#2059) requires the removal task to be filed when the flag is born: a release flag is short-lived and removed after rollout.

Problem / audience

EKS operators: until graduation, the in-place scaling path stays opt-in and the default update flow remains delete-and-recreate.

Acceptance criteria

• The in-place scaling path is exercised against a real EKS cluster (scale a managed node group up and down via cluster update; verify no recreation and correct converged sizes).
• On success: move the Updater methods onto the base EKS provisioner, delete the UpdatableProvisioner wrapper and the experimentalInPlaceUpdates spec field, and regenerate schema/CRD/docs.
• On failure: capture findings as issues and iterate behind the flag.

Rough size: S (excluding the live validation session). Part of #4328; depends on #6195.

[coderabbitai]

🔗 Related PRs

#5506 - test(autoscaler): cover the #5219 baseline-propagation routing dispatcher [closed]
#5524 - feat(scaffolder): scaffold a Dev Container on cluster init [closed]
#5525 - fix(autoscaler): grant Deployment/ResourceQuota RBAC unconditionally [closed]
#5533 - feat(workload): add --schema-location for catalog-absent CRDs [closed]

---

📝 Issue Planner

_{Check the box below or use the @coderabbitai plan command to generate an implementation plan and prompts that you can use with your favorite coding assistant.}

• Create Plan

---

🧪 Issue enrichment is currently in open beta.

You can configure auto-planning by selecting labels in the issue_enrichment configuration.

To disable automatic issue enrichment, add the following to your .coderabbit.yaml:

issue_enrichment:
  auto_enrich:
    enabled: false

💬 Have feedback or questions? Drop into our discord!

[devantler]

🤖 Generated by the Daily AI Engineer

Additional AC from #6195's review: extend diff coverage beyond scaling + instanceType (amiFamily, volumes, labels, taints, subnets, IAM are not reported by eksctl get nodegroup, so coverage needs the AWS SDK DescribeNodegroup or last-applied tracking) — or explicitly re-affirm the documented boundary at graduation. The flag's API docs state the current boundary.