[ECS][CloudWatch]: publish task Count/Status for each ECS Service

### 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**

We would like for AWS to publish task status and counts _by service_ to CloudWatch. There are two reasons for this as the existing option of using "by ClusterName, ServiceName" and selecting the `Sample count` metric has shortcomings:

* (higher priority) it is poor user experience: a user looking for "RunningTasks" by Cluster and by Service is _not_ going to click on `CPUUtilization` or `MemoryUtilization` for this data
* (higher priority) it is not accurate when dealing with periods beyond one minute granularity - when using a period of 1 minute you can see how many tasks are running during each one minute period - but when attempting to the actual number of tasks over a 5 minute period the number of running tasks needs to be divided by 5 (the number of running tasks is _generally_ between 8 and 40 as opposed to 40 and 200).

* (lower priority) it does not provide information about the actual task status in the [Amazon ECS task lifecycle
](https://docs.aws.amazon.com/AmazonECS/latest/developerguide/task-lifecycle-explanation.html) - even a subset of these status would be valuable
* (lower priority) causes issues in services that consume cloudwatch metrics (example: https://github.com/grafana/grafana/issues/15430)

**Which service(s) is this request for?**

ECS, ECS Fargate

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

I have *never* seen a client (at this point, 10 clients) have figure this out. When 0 of 10 clients can retrieve data they should use when making decisions - this is a problem.

**Are you currently working around this issue?**

User education.

**Additional context**

**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.)

The issue describes a requested AWS ECS/ECS Fargate and CloudWatch capability, but names no repository files, tests, or entry points. Start by reviewing the ECS task lifecycle and CloudWatch metric context linked in the request; done would require an AWS-defined plan for publishing per-service task counts and statuses.

Written by the indexing model from the issue text.