A serious editor makes a decision based on the reports from the reviewers along with their own critical view of both the manuscript and the reviews. This means that both reviewers may not have made the official recommendation of reject, it sometimes happens reviewers give a "major revision" officially while in their confidential comment to the editor they provide their reasoning for providing that recommendation instead of a reject, which they think would be equally appropriate. Hence there is communication "behind the scenes" that is not visible to the author(s). It is also possible that an editor makes a decision for a reject based on reviews that recommend otherwise. In such cases the grounds may, for example, be that the editor sees that the revisions will be too complex to fit the time frame of a regular "major revision". Or, that there is some formal issue that reviewers will be unable to detect. The latter should, however, not be very common since such issues, including suitability to the journal, should be weeded out at the time of submission, not after review (thereby wasting reviewers valuable time).
In the quote you provide, it seems as if both reviewers have found grounds for rejecting the paper. It is probably not very common that an editor changes such overwhelming recommendations (although it can happen). The fact that reviewers provide comments is not in any way unusual. Any serious reviewer knows that part of reviewing is to provide feedback on what is thought should be revised. Hence there is normally no major difference between reviews resulting in a recommendation for "reject" than those resulting in, say, "major revision". Any first round review that is returned without any comments and providing either an "accept" or a "reject" will, in my opinion as editor, be signalling a reviewer not willing to do the job and in fact useless for the process (manuscripts so good that they can be accepted without any action are very rare indeed).