Failures don't have to be spectacular to be useful. One of the nice things about a reusable vehicle (and one reason why rockets are so much more expensive than aircraft to develop) is that you can test things incrementally, such that a failure in whatever new thing you're trying won't doom the system as a whole.

Look at the Apollo program, for example. I doubt anyone would say they weren't trying as hard as they could, and yet the only failure I'd call "spectacular" was the very dumb and useless failure of Apollo 1. There were lots of other failures, of course. Apollo 11 had persistent trouble with the computer during landing, nearly leading to an abort. Apollo 12 got hit by lightning on launch, twice, and the mission was only saved by quick thinking. Apollo 13, in addition to the famous explosion that happened partway through the mission, lost an engine on launch. On later investigation, it turned out that the engine loss was due to massive oscillations in the rocket that would have torn it apart within a few more seconds if the engine hadn't fortuitously shut down. Apollo 14 had docking trouble, had to have its computer reprogrammed by the crew in flight to work around a faulty switch, and finally had a radar failure, all of which threatened the mission's success if not the astronauts' lives. Apollo 15's first stage didn't shut off when it was supposed to, nearly requiring a launch abort. Apollo 16 looks to be the first one that didn't encounter major problems.

I wouldn't be surprised if SpaceX didn't have a lot of similar failures, which are big deals to them but don't look like anything other than "success" to outside observers.