[EKS] [eks-pod-identity] [request]: please allow setting a custom service-account-lookup-grace-period

### Community Note

* Please vote on this issue by adding a 👍 [reaction](https://blog.github.com/2016-03-10-add-reactions-to-pull-requests-issues-and-comments/) to the original issue to help the community and maintainers prioritize this request
* Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do not help prioritize the request
* If you are interested in working on this issue or have submitted a pull request, please leave a comment

**Tell us about your request**
Please allow users to set a custom value for the eks pod identity webhook flag `service-account-lookup-grace-period`.

**Which service(s) is this request for?**
EKS, specifically the pod identity webhook.

**Tell us about the problem you're trying to solve. What are you trying to do, and why is it hard?**

We use ACK EKS custom resources for pod identity associations, and when a helm chart deploys a new service with the custom resource, service account, and deployment, the pods are often scheduled before the service account is synced. This is something kubernetes eventual consistency doesn't solve because a crashloop doesn't retrigger the mutating webhook. Despite the warning about scheduling performance when using a high value [here](https://github.com/aws/amazon-eks-pod-identity-webhook/blob/3a28f5261a026e892ad73cd51d6f5b15dd6be866/README.md), I don't really see how it's a high risk, as scheduling pods without the necessary IAM role isn't very useful anyway. As far as I understand the behavior of the webhook, it will return early as soon as the sync happens. Regardless, we'd be willing to accept the risk of slower scheduling to eliminate the missing role credential issue.

Granted this is only an issue on the first deploy of a helm chart, but we create many ephemeral environments in CI so it's a constant problem for us.

**Are you currently working around this issue?**
We're currently using helm hooks to work around this but they're fragile. We're considering installing the pod identity chart ourselves instead of as an add-on to give ourselves access to this setting.

**Additional context**
Anything else we should know?

**Attachments**
If you think you might have additional information that you'd like to include via an attachment, please do - we'll take a look. (Remember to remove any personally-identifiable information.)

Start with the linked pod identity webhook README and the `service-account-lookup-grace-period` flag described there. Trace where the EKS pod identity webhook exposes or consumes this setting; done means users can set a custom value for the webhook and the documented behavior matches the request.

Written by the indexing model from the issue text.