NavList:

Geoffrey Kolbe, you wrote:
"I would say the accuracy of estimating Time this way is probably as good as the average 'lunar'."

Depends on the meaning of "average lunar", of course. But certainly when done right and combining multiple observations during a single lunar eclipse, the resulting absolute time can be as good as one relatively mediocre lunar (just one, no averaging) observed with an average sextant. The big advantage for a lunar observed with a sextant is that you can try again tomorrow if it's cloudy. With a lunar eclipse if it's cloudy, you may have to wait two or three years. The big advantage for a lunar eclipse observation is that you can time it without a sextant!

Roger Sinnott wrote a great article that was posted last week on Sky & Telescope's website: Useful Projects for a Lunar Eclipse. There's a list of crater event times that could be used for planning purposes. 

I've been thinking about generating a light curve with a digital camera somehow. If I take an image every five or ten seconds for the duration of the eclipse with fixed camera settings, it should be possible to determine the time of mid-eclipse quite accurately. And it will make a nice video, too.

As Kelly Beatty mentions in the article Solar and Lunar Eclipses in 2019, referencing Joe Rao's article in the print edition of S&T this month, the eclipse occurs on the Sunday night before Martin Luther King Jr. Day which is a holiday throughout the US. This, along with its high altitude over the continental US, guarantees that it will be the most watched lunar eclipse in recent history (in the US at least) assuming the weather cooperates. Of course that's the big unknown. Clouds will spoil it for at least some of us. And January 21 also happens to be the coldest day of the year on average in the northern part of the continental US.

Frank Reed