While the Kill-A-Watt devices have been quite reliable, they have never really stated particulars in their specs and just claim something like .2% for everything.

Inexpensive DVMs, with frequency counters, usually fall in at around 2-3%. Then, there is also the display tolerances -which really gives you tolerances - on top of tolerances.

My understanding has always been that power companies will generally let slop occur during high demands, but then also monitor the exact frequency fluctuations, and actively tweak it during lower use times to keep the overall day at exactly 60Hz - specifically, to accommodate synchronous devices such as clocks.

I have some extremely actuate test equipment and frequency counters, and while I just never simply measured AC line frequency during the day, I think if I did, the deviations would be much less than probably 0.1 Hz during the day and average out to .01Hz overall throughout the day.

I suspect that explains what I have observed. The 60.4 is just a 0.6 % error and the 59.9 is just a little less than 0.2 % error -- much closer than 2%. For my use with the generator, I will just get it as close to 60.4 on the meter as possible.
As far as I know, all the POCO generators online at a moment will naturally stay synchronized as close as possible. The one that might lead will be "braked" by the others and the one that might lag will be "motored" by the others.

Bus driver, after thinking about this today, when I got home this evening, I did a quick check of AC line frequency using a "better quality" frequency counter which read 60.014623 Hz without allowing any "warm-up" time for it to stabilize.

I also then checked the frequency using an inexpensive DVM and it read 60.2 Hz.

I think it's safe to say the daily average will be pretty darn close to 60.01 Hz, if not even closer, after active adjustments are made by the sources that are providing the power to the grid.

This meter will be used along with an equally cheap 120 volt voltmeter in a small enclosure with cord and plug to be used with any portable generator to monitor the output as it is supplying parts of my house during power outages. Since it reads 60.4 -60.5 on the POCO power, I will use that reading as my desired target for the generator.
My interlock setup is now complete for input with up to 15KW generator. My largest generator at present is 4.4K/4KW. Input to the house is 120/240, but monitoring just the 120 tells me all I need to know.
We get water from the well. Having water and refrigeration really helps get through power outages. The small generator will also operate the propane furnace that we have for backup heat.

What kind of frequency variations would be expected from large loads going on and off line?