Edit1: Telemetry came back at 10k. For the next minute and a half there was uncertainty about the main engine even though telemetry clearly showed vertical speed going up fast. More then a minute later, at 5k a reset request was made.

Edit2: A minute goes by and at about 500m controller asked if there is a confirmation to send rest to JPL, another, announced that engine is on. Crash happens at that moment. 149m, 134.3ms vertical.

It seems like the landing program would want to have a feature in place that automatically restarted the engine in a situation like this.

I can't tell from that if normally an engine outage would automatically restart but didn't or if the condition would always require manual intervention. I hope we get more details soon. The moon seems close enough (a few seconds), but accidents always happen closest to your destination. More automation is more complexity but it will be valuable for longer trips.

Have any of these moon landings been done at night where people have been able to watch it happening through a telescope? Or are things so small at the moon's distance that there'd be nothing to see?

Actual landers tend to try to land during lunar day, to take advantage of the warmer temperatures and to have sunlight to recharge batteries.

There are lunar retroreflectors about a metre across [0] that are detectable with optical equipment, for some definition of optical. But to leave a mark that can be seen from earth - to leave your tag, or an X, or a crude drawing of a penis and testes - surely that's an urge as old as art itself.

[0]https://www.nasa.gov/mission_pages/LRO/multimedia/lroimages/...

Or with more time to recover from the outage before landing. If it happens within the last minute (wild estimation) then you have little time to recover the lost velocity.

Obviously they restarted it after this failure so it was not a complete breakdown of the engine.

I was on a plane that had one engine fail. It was fine.