gold image is a term used to describe a template for a virtual machine (VM) that contains a preconfigured operating system and applications1. A gold image can be used to create multiple session hosts that share the same configuration1. However, a gold image needs to be properly prepared before being used as a template123.
One of the steps involved in preparing a gold image is to run sysprep on the VM. Sysprep is a tool that removes system-specific data such as computer name, security identifier (SID), and driver cache from Windows23. This allows the VM to be generalized and duplicated without causing conflicts or errors23.
One of the options that can be set when running sysprep is CopyProfile. CopyProfile is a setting that determines whether the user profile settings of the built-in administrator account are copied to the default user profile2. The default user profile is used as a template for creating new user profiles on Windows2. If CopyProfile is set to True, then any changes made by the administrator account (such as desktop wallpaper, screen saver, Start menu items, etc.) will be applied to all new user profiles2.
Therefore, if users are being disconnected from their sessions consistently after 10 minutes, it could be because the gold image has some incorrect settings in the default user profile that cause session timeout or termination. To resolve this issue, one possible solution is to sysprep the gold image with CopyProfile set to True, and make sure that there are no settings in the administrator account that could affect session stability (such as power saving options or Windows Store applications)2.

A high CPU Ready time means that the vCPU is waiting too long for the pCPU, which can result in poor performance and slow responsiveness of the virtual machine (VM).
According to Nutanix best practices2, one of the factors that can affect CPU Ready time is the host power policy. The host power policy determines how aggressively the host will try to save power by reducing the frequency or voltage of the pCPUs.
The recommended host power policy for Nutanix clusters is High Performance, which disables any power saving features and ensures that the pCPUs run at their maximum frequency and voltage.
This answer seems to explain why desktops on one host are running slower and have higher CPU Ready times than on other hosts, as they may be suffering from reduced pCPU performance due to a lower host power policy.

Deduplication is a process that eliminates duplicate data blocks and reduces the amount of storage space required for virtual desktops. Deduplication can be enabled at both cluster level and container level on Nutanix.
Deduplication can also work with VCAI (View Composer Array Integration), which is a feature that uses Nutanix native NFS snapshot technology to clone VMs2. VCAI can improve performance and reduce storage consumption for linked-clone desktop pools.
Another feature that can help with capacity issue is Shadow Clones, which is a distributed caching mechanism that allows multiple VMs to read from a single vDisk copy stored locally on each node3. Shadow Clones can reduce network traffic and improve boot times for virtual desktops.
https://www.nutanix.com/support-services/training-certification/certifications/certification-details-nutanix-certified-professional-ncp-euc-v6

administrator can take to resolve the issue of performance for several full clone virtual desktops within Cluster01 by migrating them to Cluster02.
a recovery plan is a set of steps that defines how to recover guest VMs from a disaster. A recovery plan can be used to migrate VMs between Nutanix clusters at the same or different availability zones1.

This option allows you to create a custom report with the required metrics for all VMs across all clusters and send it automatically to the operations team via email4. This way, you don't have to manually export or FTP the data every month.