Ah, you're quite right, very fair point. Supersonic retropropulsion was indeed a novel technology that ended up in their critical path. Interesting to speculate on what they might've done if that hadn't worked out! Are there more off-the-shelf techniques that might've been good enough? (Eg. nose-first re-entry with a deployable heat shield.)

Anyhow, obviously SpaceX has done many other very technologically innovative things as well. But this would all have been a moot point if they'd operated per the rules of the existing industry, on cost-plus contracts with a wildly distributed workforce with each department or subcontractor focused on optimising its own particular subsystem without consideration of the whole. Breaking that paradigm was far and away the most important innovation of SpaceX today.

It's worth noting that they've now gotten the gains from escaping the industry's old local maxima, and that going forward (with the BFR) they're going to be more reliant on novel technologies than they have been to date. But they'll also be reliant on the emergence of new markets for spaceflight at the much-lower price point. I suspect that StarLink is part of this effort -- apart from being a good idea in its own right, it's a way of Musk saying to the world: "See guys? This kind of thing is what you can do with really cheap and reliable spaceflight! Start thinking on this scale, people!!!"