In assembling the custom list of Grand Canyon National Park summits, I found that the location used in the LOJ database for the GCNP highpoint is not actually in the park. The cohp page lists 4 possible areas for the park highpoint. The LOJ location is within "area 4" but lies north of the park boundary. Roach pinpoints a location within "area 2." peakbagger.com uses the lookout tower location within "area 1." Ken Jones' report there indicates that he thought "area 2" was higher than "area 1."
I think John likes to use the "largest" contour when there are multiple possibilities (?), which is perhaps why "area 4" was used in LOJ. Looking at the 4 possible areas, "area 2" actually has the largest area within the park boundary, so that might be a better choice for the park highpoint location.