Track records are tricky for rocket systems. Since the first Sputnik, over the 57 years the humanity had may be - roughly - 4000 launches of space rockets of all kinds, with many different fuels, engines, sizes etc. It's relatively rare when you have a good track record which is measured by many tens of flights - let alone hundreds. When a system flew 20 times it's already considered rather mature - double so if those were all successful flights. So it's quite usual to use a new engine in a new design of rocket - in fact it's rare when you have many different rockets with the same engine type or many different engines used on the same rocket. Rockets and engines are usually optimized to each other - particularly rockets are built around engines, since margins are quite thin.
It doesn't matter how reliable the engine is if it's not performant enough for the envisioned task.
As to your question about reliability and track record, it's kinda hard to base the decision on past launch performance since most rockets-engines are built to a specific rocket and vice versa. These things aren't like a used car where you can look at thousands of unique datapoints, or use the same engine in the silverado that you use in the corvette (yeah, this is actually true. different ecu settings, though).
And it generally cost prohibitive to do a full dry-run with only a cargo full of, say, bricks, because a single launch costs well north of a hundred million dollars. Instead you have to run a lot of ground based tests and figure a bunch of probabilities.
These guys won a one point something billion dollar contract from NASA to deliver shit to the ISS. NASA's not exactly well funded and not exactly known for flying by the seat of their pants, even by government agency standards. In order to win that contract they needed to prove to NASA's satisfaction that there was a high probability that their rocket would get to ISS without a major problem, let alone one as catastrophic and uncorrectable as exploding a few hundred feet above the pad.
It doesn't matter how reliable the engine is if it's not performant enough for the envisioned task.
As to your question about reliability and track record, it's kinda hard to base the decision on past launch performance since most rockets-engines are built to a specific rocket and vice versa. These things aren't like a used car where you can look at thousands of unique datapoints.
And it generally cost prohibitive to do a full dry-run with only a cargo full of, say, bricks, because a single launch costs well north of a hundred million dollars. Instead you have to run a lot of ground based tests and figure a bunch of probabilities.
These guys won a one point something billion dollar contract from NASA to deliver shit to the ISS. NASA's not exactly well funded and not exactly known for flying by the seat of their pants, even by government agency standards. In order to win that contract they needed to prove to NASA's satisfaction that there was a high probability that their rocket would get to ISS without a major problem, let alone one as catastrophic and uncorrectable as exploding a few hundred feet above the pad.