This sounds like a survivorship bias, you imagine that people who blow stops won't do that if you don't stop but, people who did the same on a bike and were t-boned by a car might not be in a good condition to post about their experience anymore.
Anyways, even if you are 100% right and not stopping at stop on a bike will prevent people blowing stop completely (and people who have the right of way will also yield to your bike because they will see you 100%) how is it safer than to stop and proceed when there is no traffic across?
Behavior by cyclists certainly has a bias towards survivorship. Every other vehicle on the road represents a potentially lethal threat at any time.
Inadequate dedicated infrastructure for cyclists leads to behavior like "Idaho stops" that look counterintuitive to drivers, but improves safety for cyclists at intersections.
>how is it safer than to stop and proceed when there is no traffic across?
Generally, on a bike you want to be predictable. If you do something weird, like not going when it's your turn, you're increasing your risk. And in many situations if you wait for there to be no cars at a stop sign, you're going to end up holding up cars behind you and be on the receiving end of road rage, or drivers behaving recklessly to get around you. Plus you may be waiting literally hours for all the traffic to dissipate, and I prefer to be home before sunset because riding at night introduces other safety issues.
Anecdotally I've never felt at risk of being t-boned when doing an Idaho stop, but it happens regularly when I do a complete stop (in part because it's unpredictable - drivers don't come to a complete stop at the line so they misbehave and don't know how to respond when they see a cyclist stop).
Beyond my anecdata, there's plenty of more rigorous studies showing the benefits of the Idaho stop.