If the question is "Should the author publish the reviews he got for his paper(s)?", then I'd say no (maybe with some exceptions, but this would have require a per-case discussion). I believe the other answers have more or less covered that.
But, if the question is "Should the journals publish the reviews?", then I believe it would be both ethical and useful to publish the final positive reviews, and I have several reasons for this.
First, giving a positive review is like giving a positive grade on an exam. The person doing so should stay behind his "verdict" with his professional reputation. The review is one of the results of the work that the researchers do, so publishing positive reviews doesn't seem to me much different from publishing the results of the research in papers.
Second, and quite related to the first, I've read some really crappy papers, with nonsense, obvious errors, misquotes, etc. Reviewer cannot "catch" everything, but some of the papers get bad enough that it is obvious that the reviewer didn't do his job. If the reviewers knew their names would forever be publicly associated with such paper, I believe some of these might actually try to do their job.
As for the negative reviews, I see no point in "shaming" the author if his paper was too bad (in whatever way) to be published. If this was not the case, but the reviewer is to be "blamed" (i.e., for misunderstanding the paper), the issue can be resolved with the editor, or the paper can be submitted elsewhere, again giving no reason to make the negative review public.
One might argue that the negative review is also like publishing the results of the research in papers, but I see it more like a failed research, which is not something that one usually publishes.