Maybe it's just little old me, but does anyone else get the impression there are serious issues with the SQL Server instance itself (or instances as the case may be) or possibly the whole infrastructure?
For example, I'm really having a hard time believing (vis-a-vis Gail's response) that simply connecting to production via SSMS is more likely to "bring down the server" than using RDP to accomplish the same thing or that using a Transaction can "destroy" a server.
No doubt a lot of issues are self-inflicted due to these rules, but the "reasons" given for them seem a bit excessive under typical conditions.
I suspect that it's a question of unreasoning responses to problems. For example, someone opened a transaction, modified a substantial portion of a database and then didn't have a commit which lead to massive amounts of blocking. This appears to be a "destroyed" server since everything is stopped. The answer that they came up with, never use transactions. When, the answer should be, use transactions appropriately. Same thing with SSMS. I've seen people run queries from SSMS that hurt the server, and if lots of people are all running queries through SSMS that hurt the server, then it just makes sense, in a twisted way, to stop using SSMS, or least focus it through a single RDP point so only one bad query can run against the server at a time.
It's just logical conclusions based on incomplete and incorrect information. It happens all the time.