I don't believe the Amateur photographer figures. I am not sure how they are testing and calculating them. However I do understand how DXOmark is calculating them and I have an understanding how Bill Claff computes his numbers. I have used the method proposed by DSPographer (
http://www.sensorgen.info/Calculations.html for those so inclined) to calculate the K5 and K3 but I haven't applied Bill's method to calculate the KP and K70. They are a bit more involved. There are a number of things to keep in mind when discussing DR. First between what levels are you measuring, particularly what is the acceptable lowest signal? And what is the ISO value (the camera value as used by the manufacturer or some calculated value) used as the independent variable (x axis)? As you can see by DXOmark some manufacturers use pretty strange ISO values. And are we talking about a DR for normalized output (like DXOmarks "Print") or DPReviews "comp" function)?
But be that as it may, to the questions asked. The K5 has the same DR as the K3II when normalized to the same output size. Non normalized DR (not accounting for the number of pixels) the K5 is about 0.7 of a stop better at about ISO 100 (measured value). I expect, but haven't computed it, that the K70 and KP would be better than the K3II by about that .7 of a stop at ISO 100 (calculated). See the file marked DPReview2 below. They'd be about where the 7200 is at abase ISO and likely better than that as ISO increases. See the file marked DPReview4 below. The KP seems to be something like the sensor used in the Sony 6300 and 6500, the electronics may be a bit better in the camera. See the file marked DPReview3 below. The noise is pretty well controlled and the KP and K70 look like they have less noise to my eye that the 6300's. If its the same sensor tech we might have the same or close to the same Full Well Saturation. So the DR would be in that same ballpark although it might be a bit better particularly as ISO increases.
All this to say is that the KP is broadly on the same performance as the Nikon 7200 and Sony 6500 (using raw only - the JPEG engines of either are better tuned than Pentax's). So much so that you'd likely not notice the difference. Would you notice the difference between the K3II and the KP? Everything equal and if you are really pushing the envelope - maybe - about the same difference in performance from say a 7100 to a 7200 which is to say pretty small (the biggest reason 7100 owners upgraded was due to banding issues which the K3, K70 and KP do not suffer from). The KP seems to test better (by the comparison photos) but other features about the K3II may be more desirable than those of the KP.
On a personal note I have a K3 and a K5. I don't feel the need to upgrade to the KP. The K3II replacement maybe. I have some lenses I want first that would make a bigger difference to my photography.