The previous discussion on this is here: https://discuss.okfn.org/t/election-data-criteria/1597
I infer that OKI's point of view is: 'the optimum balance between openness and privacy is to report vote counts down to polling station level'. Although there is no research in this area to back this up, the results of the survey suggest that about a third of countries do publish to this level, so there is some justification.
My main problem with the methodology is why do OKI decide that 0% is a representative score, when there is still plenty of value in the data which is published in this area. Countries who have have published no result data are getting the same score as UK.
OKI knocks off a percentage point for not being up to date or providing bulk download, but other minor issues like level of detail knock off 10% of a final score. If there is no nuance in the scoring then the final rankings don't mean much.