The HP and Rosewill remotes designed for WMC (Windows Media Center) both worked the same way. After matching commands to Flirc, they worked but - not really. I eventually figured out that approximately only every other keypress did anything. After further research, this thread included I figured out the issue seemed to be the toggle bit.
At one point, I reached the maximum memory of the Flirc so I wasn't too thrilled with the idea of programming each button twice and so I never gave that a good try. I may try again if I find some time - no promises.
Would I need to go through the entire button programming sequence twice? Like:
Press play on Flirc GUI
Press Play on MCE remote
Press play on Flirc GUI
Press Play on MCE remote
?
I can see how that might work but it sure seems tedious and also would it take up double the normal memory? It would be much nicer if you could somehow use the desktop application (gui) to decode the remote protocol and assist with the Flirc programming. It could then recognize the RC6 protocol and prompt the user to press the key two additional times, for example.
I admit after reading a few of the protocol specs, my head is totally spinning. I didn't anticipate these things would be so complicated.
One more thing - I checked the modulation frequency on both my MCE remotes again. I came up with 36 kHz again. I used the DSO Nano - I have to set the timebase so that I just see the very beginning of the command and also scale and offset the voltage (essentially zooming way in) but then the ripple is extremely easy to see.
It doesn't matter to me personally what priority you put on something like this. I am sure it will help the project in the long run but for me I want results very soon so I am trying a combination of Sony remotes and Harmony 300.
I do like the Flirc concept a lot and that is why I was willing to abandon the MCE remotes in the short term.