I am looking for the same thing for my environment. Not just the UnDP but also the process of setting up correct alerting to determine a failure of any hardware which is not reported by the OS.
Outside of the shared UnDPs, the best way is to walk the MIBs on the device to see what it will support. There’s a MIB Walk tool in the engineers toolset. If you don’t own it, you can use it in the eval copy for 30 days.
To set up the alert, you will need to define both the UnDP name, and then the value that you want to trigger as the alert actions.
Like this: http://screencast.com/t/YmQyNjdlNz
Here is set of UnDP (Universal Device Poller) configurations that were uploaded to the Content Exchange that are designed to get specific HP Insight information from the Proliant line of systems. We are also an HP shop and these work well out of the box or you can tweak them as necessary.
HP Proliant Pollers:
As suggested by a previous poster; doing a MIB Walk on the system with the Engineering Toolset is a great way to see what data is available to create UnDP's for.
After you have the pollers all created you will need to create all the necessary alerts.
What you are trying to do is all possible, it will just require a bit of work on your end.
Hope this helps!