Cluster Validation Report improvements
From my experience the success of the cluster validation may depend on the network between the Failover Clustering Console and the actual nodes. In my case Active Directory Configuration Validation was failing (while everything else was successful) because of the WAN accelerator messing up communication between the machine with the console and the validating cluster nodes (which were being validated across the WAN). Besides, I also noticed that the longer the latency between the console machine and cluster nodes, the longer it takes to deploy a cluster. If these tests could be refactored to be executed locally on each individual node and then report back the states through a common protocol (i.e. through WS-MAN) the success of the report generation process wouldn't really depend on the validator party.
Ivan Seriavin commented
That could also open up a scenario when a cluster validation can be triggered from a Failover Clustering console running on the OS of an older version. For example, currently, in order to deploy Windows Server 2012 R2 clusters, it's not enough to upgrade SCVMM, but also its underlying OS.