I'll try that...
I did some quick tests with my K-5 and my old K20.
I was originally disappointed with my K20's PTTL performance until this thread came out -
https://www.pentaxforums.com/forums/pentax-flashes-lighting-technique/103779-...x-flashes.html - Using the trimpot, I was able to get the K20 to perform well in nearly all situations - bounce and direct, wireless and no wireless.
I did these tests at ISO 800, f/5.6, 1/60 second, 70mm on a Sigma 17-70. In my case, the flash was zoomed out to 16mm without any diffuser (to get a wide diffuse light source out of my white ceiling)
When shooting a whitish-brown carpet, I expect the camera to underexpose and bring the carpet to around midway on the histogram (e.g. expose the scene to 15-18% grey) - When dialed in with the trimpot, as I stated above, the K20 puts the spike on the histogram at just above dead center (I ran my K20 + AF540 a bit hot since I have -2 stops FEC and only +1). It does this whether bouncing off the ceiling or direct, and whether wireless or not.
With the K-5 and the flash dialed in the same way, same ISO, focal length, and aperture:
Direct on-camera flash is a little bit overexposed (maybe half a stop?) compared to the K20
Bounced on-camera flash is significantly overexposed (maybe 1.5 stops?) compared to the K20
Wireless off-camera flash is significantly underexposed (about 1-1.5 stops) compared to the K20 - whether direct or indirect
Edit: The trimpot affects on-camera performance but not off-camera - wireless P-TTL always underexposes severely
I'll try to get some samples later this week - my arm got tired!