This totally makes sense. If the paper shows that A correlates with B, and interprets this as A causes B, but doesn't show it; and if you think B causes A; then this is a great explanation for their result.