You hit at exactly the crux of the matter. It seems that for whatever the reason, some of the judges chose to define "influential" as essentially being synonymous with "recognizability" and "popularity".
But that isn't what influence is all at. As you alluded to, "influential" is one who influences, and influence is essentially causing change, so the measurement of influence is in the amount of change caused. Measuring change requires taking a number of things into consideration, but what would almost always carry the most weight in that measurement would be the number of people affected by said change.
So essentially the list was billed as and intended to be the "The 15 people who caused the most change that affected the most people in 2020". Yet when we all looked at their top 15, it largely doesn't represent this at all. A number of the people in the top 15, even high in the top 15, were responsible for causing little change among few people. Other people that were clearly responsible for causing notable change affecting large numbers of people were noticeably absent from the list completely. It is clear that the quality of the judging, or of the methodology used to "score" the voting, or both, failed miserably at what they were supposed to be doing and what they claimed to be doing.
And now it appears that that the person responsible for choosing the judges, and for choosing the judging methodology, and for actually doing the final "scoring" which apparently was done in private, is the agent representing several of the parties on the list who were among the most out of place and least deserving of being there. This leads to the possible impression that the failure of the top 15 list to deliver on what was intended and promised may not have been as much due to incompetence as perhaps being due to having been intentionally orchestrated to meet other motivations of the list organizer that were far less honest and sincere.