I agree that they have to do better. That cloud was likely a flameout that blew methane out before termination when the oxygen ran low.

I consider that these tests have to have some tolerance for failure as the finished product shouldn’t result in any venting except in an accident.

The alternatives like keralox, solid fuel or hypergolics have worse emissions in actual operation and hydralox wastes vast amounts of energy refining and chilling liquid hydrogen, so getting the methalox cycle running will be a net benefit once testing is complete.