Based on what data? I've found them to be as accurate or more accurate (and the data I've seen says the same) except around sleep tracking, where Garmin is worse. But it's not hard to create a function to correct for that if you really care about it, it is good enough unless you need vanity metrics

That's an odd objection. Garmin's optical heart rate sensor accuracy might be slightly worse than Apple's under certain limited conditions, depending on how you test. But anyone who really cares about precise heart rate uses a chest strap anyway.