[aws][compute] research changes to use by default nitro instances (m5*) on IPI

USER STORY:

As a SPLAT engineer I would like our customers to have a good experience running the quick installation on AWS (default IPI config) using the current generation of EC2 (Nitro Instances), which also has a lower price and better performance.

DESCRIPTION:

AWS launched Nitro instances in 2018 with significant improvements on hypervisors, making it possible to create new instance types faster (as we can see the variations on v5). The default instance type of installer is m4.xlarge (build version[1]), and the m5.xlarge is valid and can be changed through install-config customization.

The Nitro instance has also enabled by default the EBS optimization [2] that can increase the performance of volumes with no additional costs.

ACCEPTANCE CRITERIA:

• 1) BZ or PR on openshift-installer changing the default EC2 to m5.*large : control plane and compute node
• 2) Review the documentation for any reference to current instance type (m4.*large)
• 3) Map the changes to be done to enable the following flags when the instance is provisioned: EBS Optimization

ENGINEERING DETAILS:

[1] installer build version (from master on 20210810)
~~~
$ bin/openshift-install version
bin/openshift-install unreleased-master-4909-g7f8dfaf0ea9fe97d82fb2f4f1eac21bfe694aba3
built from commit 7f8dfaf0ea9fe97d82fb2f4f1eac21bfe694aba3
release image registry.ci.openshift.org/origin/release:4.8
~~~

[2] Review: How can I get maximum I/O performance from my EBS volumes hosted on EC2 Nitro-based instances?

[3] AWS API run instance options