Regarding Disaster recovery testing, I am looking at how other companies do disaster recovery testing in order to benchmark. Do they provide isolated environments for their teams. Do they have a centralized team that guides them through this? Do they recover from backups or rebuild using IaaC. Do they recover in a different region? Does anyone test recovering all apps in an Azure/AWS region in a second region or do they recover all their applications on-prem in a second physical region?

Thanks for the question. I would like to share my company experience on this:
1) We identified around 20 critical applications that the business needs in case of disaster. Only these applications have an environment in the DR site which is a special physical place.
2) We also identified the positions (people) who are required to be in DR situation and the equipment they require.
3) The DR test is done once annually. The process is to notify all staff that this day will be a DR drill. The main Data Center along with the high availability site are shutdown. This should trigger the cold DR site to start. Services such as Domain Controller, DNS and foundational services are moved top the DR.
4) The people identified as critical will move to the DR offices and start working. The test involves accessing all the critical applications. The working environment is simulated for 4 hours. After which the DR site is deactivated and the original sites are brought up.
5) The main learnings are captured and corrected if needed. For instance, some cases prompted the need for creating a fall back internet route from another provider. This was captured for the next DR drill.
6) The DR drill involves the simulation of a real crises, thus critical infrastructure like police, fire, hospitals, ..etc. are involved to ensure readiness.
Hope the above helps.

In our organisation we are providing different environments for disaster recovery testing