chore(lambda-runtime): slightly optimize graceful shutdown helper by using new tokio::try_join biased; api

[jlizen]

✍️ Description of changes:
Closing out some minor to-dos. My tokio PR adding a biased; mode to join/try_join landed. This allows us to skip the bookkeeping of keeping track of which branch of a join/try_join! statement was polled last, and always just poll in order.

This is a minor optimization for our graceful shutdown handler, since we always want to just poll our graceful shutdown future first before our non-op extension. So we can save a bit of memory on the stack by skipping storing a usize. It also very marginally could impact latency (by micros if not nanos, it's insignificant probably).

Also applying a similar optimization to the internal extension example where we have it the extension wait until after the main handler completes so that it can flush telemetry.

This involved explicitly setting tokio to 1.46. This won't impact our MSRV as our MSRV is set to 1.81 but tokio is 1.70.

🔏 By submitting this pull request

• I confirm that I've ran cargo +nightly fmt.
• I confirm that I've ran cargo clippy --fix.
• I confirm that I've made a best effort attempt to update all relevant documentation.
• I confirm that my contribution is made under the terms of the Apache 2.0 license.

[jlizen]

Clippy check is failing due to unlined format args changing to warn level.

But I have this fixed in: #1005

Probably fine to ignore in this CR in favor of that fix.