It‘s definitely possible and happens regularly. But statistically speaking, it‘s a freak accident compared to car vs. pedestrian. It‘s orders of magnitude less likely.
What is your point then, if not “cycling drunk is as likely to kill somebody else as driving drunk?” If it is “it’s technically possible to kill somebody cycling drunk” - then yes, I concede the point. But it’s not a useful point in this discussion.
