Just as a word of caution: a computer model is just a model. If there are enough parameters you can tune, it’s fun to play with, seeing what can happen. But just because you get a result from a model doesn’t mean it’s guaranteed to happen. Models are written by people, they have software bugs, and more. Obviously they wouldn’t publish this unless they were confident, but the scientific method is a feedback loop, and we’ve only been through that loop once here so far.
What needs to happen now is other scientists need to recreate the model. If it turns out that multiple models point this way, then it’s time to panic.