You listed categories of technology. Good start. For stacks think racks, then chassis, then server hardware, server OS, middleware including NIC and RDBS, etc. and then the business application. Please note that the server will have infrastructure applications like anti virus, OS agent, hardware agent, RDBMS agents, etc.

Now for your standard change, think about repetitive installs or updates. So security patches may be the first thing since they are mandatory and most often with the least disruption. Then think hardware upgrades like NIC cards, memory or additional hard drives. Some these may be non-disruptive because power is not disrupted and no restart is required. Others require a power down. But the common thread is fast and easy installs with know results. You do it all the time to lots of servers or devices. You will do it all the time. Get these common installs sets up as the first wave of standard changes so the overhead of approvals is reduced and time to market is greatly reduced. Once this set is working well, then you can expand what is a standard change because you have a track record of success. The idea is to have normal changes be things that are not common. Therefore, these not common changes need more reviews and testing to protect the production environment from incidents due to change. If you have not done something enough for it to be common or routine, then keep it a normal change.