docs: the comparison covers Temporal and Cloud Tasks, not just Celery

docs/content/docs/about/comparison.mdx compares FlexiQ to Celery, and the footer links it as
"Celery comparison". For 1.0 that was the right frame.

It is now wrong in both directions. It overstates the Python framing — FlexiQ is one engine
under three SDKs and a network door — and it understates the competition, because the tools a
team actually chooses between post-2.0 are not all Python task queues.

Add, at minimum:

• Temporal. The closest comparison and the one most worth being fair about. Both do durable
  execution with memoized steps. Temporal has a far richer workflow model, versioning, a mature
  UI and a hosted offering. FlexiQ is one binary with no cluster, runs on SQLite, and is a task
  queue first. Someone evaluating both should be able to tell from the table which one they
  want, and for a large class of users that will be Temporal. Say so.
• Cloud Tasks / EventBridge. The managed alternative. Zero operations, no local development
  story, no durable steps, and a dead-letter queue you cannot replay from.
• BullMQ. The Node incumbent, and the comparison the Node SDK is actually measured against.

Keep Celery, Dramatiq and RQ. Rename the page and the footer link so it is a comparison, not a
Celery comparison.

Every row needs the benchmark (#824) behind it or no number at all. A comparison table with
unsourced performance claims discredits the honest rows next to it.

Start with docs/content/docs/about/comparison.mdx and locate the footer link that calls it “Celery comparison”; also read benchmark #824 before adding any performance row. Expand the table to cover Temporal, Cloud Tasks/EventBridge, and BullMQ while retaining the existing queues, then rename the page and footer link and ensure every number is sourced or omitted.