I remember there was a phase of Falcon design where it looked like they had perfected barge landing, and then they had a rash of failures. Later on they admitted to intentionally crashing older boosters so they could find the limits of the hardware. They were iterating at such a pace that the data was worth more than a recovered booster. I wonder if that was the case today?
They demonstrated the engines re-lighting in space, which is significant. There had been some questions about this because the engine is of a design that is said to be very tricky to start, and the tank pressurization system of the rocket has the risk of water and CO2 ice forming in the methane tanks, which had caused several failures in past tests flights. So this is a pretty good milestone.
So we might start seeing test flights actually entering orbit soon. Possibly even carrying some real payloads soon.
Does an ocean landing cause significant damage that's not present with an on-land chopsticks landing?
Presumably there are pretty big advantages considering how much it must have cost to develop the chopsticks approach.