After reading some comments from the community, it appears that this 33% are overestimated.

• Indeed, many boosting (via the use of points) are not carried out. And, this for articles a priori very interesting.
• The Question: Can Ecency keep its promises for the Bonusvoting system ?