EASA Directive for Trent 900

The EASA found an oil fire in the HP/IP turbine area to be the cause of the uncontained failure in the #2 Trent 900 engine of the Qantas A380 on Nov. 4. This is an outcome that was expected by experts. The AD dated November 10, 2010 can be found here.
But why does an oil fire lead the IP turbine disk to disintegrate? Let's talk about what (likely?, maybe?) happened in detail and step by step:

1. The buffer cavity between the HP and the IP gets contaminated with oil. Why this happened is pure speculation at the moment. It could have been a bearing that leaked, it could have been oil lines that leaked or failed or something else...
2. During the take-off and climb phase temperatures in the cavity are rising high enough to let the oil ignite.
3. The IP shaft gets overheated and fails.
4. The IPT disk, now running free without the IP compressor as a brake, accelerates in milliseconds as fuel is still pumped into the combustor. The bore of the disk is designed to withstand speeds that are up to, say, 40% higher than the speed at takeoff, whereas the airfoils are designed to fail earlier, at maybe 120% speed. But the oil fire also heated up the bore, decreasing the so called overspeed burst margin.
5. The disk burst at a speed before the airfoils fail.

So what needs RR to do to prevent such a failure to happen again?

First of all, prevent the oil leakage. Without oil leakage no fire, no shaft shear, no disk burst.
Secondly, as you probably never can eliminate the possibility of an oil leak for 100% sure, a functionality has to be build into the engine that detects a shaft shear and shuts off the fuel immediately. Without the fuel being burned and the energy put into the turbine gas path, the disk will not accelerate.

Rolls Royce said that the Trent 900 and the Trent 1000 uncontained failures are unrelated. That might be true with respect to the causes of the two incidents. But I would like to hear what Rolls Royce changed in their design philolosophy between the Trent 900 and the Trent 1000 (and the Trent XWB) to prevent such an overspeed beyond burst limit speeds in case of a shaft shear as it appears that the outcome of the different causes was the same.
To clarify, the story I pictured is my view on what might happened. I can't assure that it was like I wrote. I hope to hear the "real" story tomorrow from RR.