[ECS] [request]: Start NLB health checks without delay

**Tell us about your request**
Newly created ECS tasks are registered with the target group, but sit idle in the NLB target group with state "initial" for 70-120 seconds before the first health check comes in from the NLB. Please reduce this time to zero, so that the NLB health checks start right as the target group target is registered.

**Which service(s) is this request for?**
Fargate, ECS, NLB (and CodeDeploy with ECS blue/green deploy)

**Tell us about the problem you're trying to solve. What are you trying to do, and why is it hard?**
I'd like for CodeDeploy ECS blue/green deploys (with LB type NLB) to not take as long as they currently do.

Additionally, during a blue/green deploy, a CodeDeploy AppSpec AfterAllowTestTraffic hook can call a Lambda to verify the new task set, but there's a danger there where connecting to the NLB test listener will send requests to the new targets (in state "initial") because the NLB has no healthy targets (and thus distributes connections to all targets in the target group). This can make the test pass prematurely and the production listener rule switches over to the target group, which still does not have any healthy targets. This would likely still work, but seems unclean. The AfterAllowTestTraffic Lambda should therefore also check target health of the new task set.

**Are you currently working around this issue?**
No workaround for now.

The request concerns Fargate, ECS, NLB target registration and CodeDeploy ECS blue/green deployment, but names no repository files, tests, or implementation entry points. Start by investigating the AWS service behavior and documented health-check timing; done would require an accepted way to remove the initial delay and keep the deployment health validation safe.

Written by the indexing model from the issue text.