The specific activity of K-40 is 0.0000071 Ci/g.
The Plexus site in particular correctly cites the 8.54E-10 figure "per gram of natural K", but also cites it as shown below (stated units = curies/gram):
http://www.iem-inc.com/information/tools/specific-activities
Again, either the ANL booklet has a typo or there's a more fundamental problem (as published, it is either right or wrong...I lean toward a typo); a better source for the radioactivity of Potassium 40 exists...somewhere. Shouldn't it?
OK, no argument there.
but also cites it as shown below (stated units = curies/gram):
http://www.iem-inc.com/information/tools/specific-activities
Yes. That gives the radioactivity of K-40 as 7.04E-6 Ci/g which is slightly off the ANL value of 7.1E-6 Ci/g, and the radioactivity of natural K of 8.54E-10 Ci/g (854 pCi/g). The second line merely indicates that K-40 is the only radioactive isotope of potassium found in natural potassium, unlike most of the other elements in the table which have multiple radioactive isotopes. There are many other radioactive isotopes of potassium, but they all have a half-life of less than a day so won't be found in meaningful quantities in natural potassium.
Again, either the ANL booklet has a typo or there's a more fundamental problem (as published, it is either right or wrong...I lean toward a typo); a better source for the radioactivity of Potassium 40 exists...somewhere. Shouldn't it?
I guess I'm just not seeing your problem. K-40's radiation is the same in all the sources allowing for slight roundoff or measurement accuracy errors. Call it 0.0000071 Ci/g, 7.04E-6 Ci/g or 7.1 µCi/g you have the same result. Dilute that with 99.988% non-radioactive isotopes of potassium and you get 8.54E-10 Ci/g ( or 854 pCi/g ) of the natural mixture. All the sources on the net have about the same value, even backing out the value from the farcical "banana equivalent dose" which largely comes from the potassium content of the standard banana.