My guess was there was an attempt to keep two engines running at landing, but one of them flamed out just after they shut down the one.
To me the SN10 flight was surprisingly soon after the SN9 failure to have the whole landing sequence redone (first ignite 3, then if successful, shut one down).
To me it looked more like there were two groups of engineers, each group preferring one way (light two vs light 3 and shut down one), arguing about how important the advantages/disadvantages of each approach are, so they designed in actually both scenarios (having no redundancy in case of SN9 vs handling the 4 liquid hammer surges in the fuel piping in the SN10 would be my guesses for the disadvantages). They decided to try first the "light 2" and then later the "light 3, shut down 1".
Well, both have failed so far, so they have some homework to do...