Outlook 2007 - Retrieving Data

We have a few of users (5-10 from 300) that report Outlook Hanging/Freezing and receiving the Outlook is retrieving data from exchangeserver.fqdn message on regular occasions every 5-6 weeks. Whilst the Exchange servers or Domain Controllers don't appear to be suffering from any performance issues (according to the VMWare Perf Stats); we have found that if the Exchange Servers (CAS/Mailbox Dual Role) are rebooted the issue doesn't occur for another 5-6 Weeks. Rather than continually rebooting our Exchange Server we want to determine the cause.

Having reviewed many KB docs on potential causes and following the Exchange Team Bloghttp://blogs.technet.com/b/exchange/archive/2005/05/25/405353.aspx I haven’t found a root cause yet. Wasn't able to get PFDAVAdmin to export mailbox item counts for affected mailboxes (so if anyone knows any other tools to achieve the same thing let me know), as I’m one of the affected users I know my mailbox isn’t huge (420 MB and whilst I do have quite a few mails I don’t think I breach the 2500+ Items per folder rule).
So I thought that rather than going direct to PSS (£££) with the issue I would see if anyone had experienced something similar in the forums first.

Our Environment 

We are running our AD/Exchange Environment on 2 x DELL R815 Servers (48 Core CPU's, 128 GB RAM) running VMware ESXi 5.1.  Our W2K8/Exchange 2010 Environment consists of a W2K8 R2 Domain, FFL = W2K8R2, DFL = W2K8R2. The domain has 1 Root DC (1 CPU, 4 GB RAM) and 1 Child Domain (with 2 DC's each with 1 CPU and 4 GB RAM).
All 3 DC's exist on the same IP Subnet and within the AD environment existing within the same site SITEA (not default Site). The IP Subnet is defined within AD and associated with SITE. All network connections (multiple NIC’s from each ESXi Server) are running @ 1 GB into the same switching environment.  

The Exchange Environment is made up of 3 Servers (1 Hub Server - 1 CPU, 4 GB RAM and 2 x CAS/Mailbox Servers – 4 CPU and 12 GB RAM) all servers are running Exchange 2010 SP2 (Build 14.2.247.5) and also exist on the same IP Subnet as the DC’s, again all network connections (multiple NIC’s from each ESXi Server) are running @ 1 GB into the same switching environment.

The Workstations are currently running XP SP3/Office 2007 SP3 and also have 1 GB Network uplinks into a separate network switching environment and are linked via a routing LAN and F/W to the Exchange/AD Network

We have run BPA on the environment and apart from reporting no writable DC’s in Default Site (fixed in a Roll Update) the only other issues reported is MaxPageSize (LDAP Queries) Value on Mailbox Servers is 2000 rather than the default 1000.
Whilst I haven’t reviewed the release documents with SP3 or the roll-ups that are post SP2 so this issue is potentially fixed I wondered if anyone had seen the same issue and resolved it.