The question should be posted in a different forum, because it has nothing to do with Integration services. As for your questions:

1) You can configure each group which node can be it's owner and also which node is the preferred owner.

2) You can check the value of serverproperty(' ComputerNamePhysicalNetBIOS'). Each time that it changes, you had a failover.

3) You can find the answer here – http://blogs.msdn.com/b/cindygross/archive/2009/06/10/sql-server-2005-clustering-tips-references.aspx%5B/URL%5D

4) Number of concurrent users has nothing to do with clustering. It is determined by SQL Server's configuration option – " user connections".

5) You can access any SQL Server (clustered and none clustered) by its name or by its IP. When you have a clustered SQL Server, you don't know on which node it will be at any time. If you'll use in the connection string the real name or real IP of the server that holds the SQL server at that time, after failover you'll have to modify all the connection strings. In order to avoid that, you have to work with virtual name or virtual IP that will get you to the cluster service which will send the connection to the active server. Virtual name is the cluster's name and the virtual IP is the cluster's IP.

Adi