IMO using things like [likely] has more potential to shoot oneself in the foot than improve performance.

Using real execution statistic to determine branch prediction (at compile time) should be the new standard