If you subscribe to the theory that people have some amount of maximum hardship they're willing to undergo for the environment, then we should focus on maximum impact, minimum hardship actions. Things like "no meat", then, which have significant hardship and insignificant benefits, displace other, more beneficial changes. A similar argument could be made for plastic straw bans.

Basically - if the choice isn't "nothing OR stop eating meat" + "nothing OR switch to more fuel efficient cars" but instead "nothing OR stop eating meat OR switch to more fuel efficient cars", then encouraging "stop eating meat" and not an actually significant change is suboptimal.