Absolutely! Anyone can comment on anything, as long as they exceed some small nominal reputation.
Restricting comments to experts, that's the huge mistake all the other sites are making. This is the opportunity for an open-review site to succeed big. This is why I am optimistic about the future of this site, because it is dedicated to high quality physics, and to openness and no restrictions using authority or power.
Just because a person studies quantum mechanics does not mean that they can't see the mistake in a fluid dynamic paper immediately. It does not mean that they aren't capable of seeing that another author's attempt to claim that a superconductivity paper is derivative is simply an attempt to pad the resume. The explanations are allowed to be lengthy, and if you have someone who makes a mistake, you can explain the mistake, and there is no need for an authority mechanism at all.
In the refereeing section, you informally encourage people to vote sincerely. That means, vote only when they think they understand what's right and what's wrong pretty well. Even still, they will make mistakes, but sincere mistakes are easy to fix, you simply point out the errors, and the community self-adjusts.
It is important to TRUST online communities, they have a good idea of what's going on, much more so than any group of experts. Frankly, the main reason is that people who don't know anything just sit on the sidelines and don't vote. But the other reason is that there are people who are capable of commenting from outside the power structure of a given field, who can demolish low quality work, but inside the field, the experts don't dare, simply because the low quality work comes from a person with power.
The way to deal with incompetent voting (only if it becomes a problem) is simply to request that the donwvotes/upvotes be specific in their claims, so that a downvote on accuracy comes with a specific mistake pointed out (usually by upvoting someone else's review which points out the mistake), and that an upvote with a sincere appreciation of the utility and scope of the paper, usually by upvoting some extension, or by pointing out some important derivative work, or simply by aesthetic appreciation of a masterful physical insight. The people who vote by definition are always the people who care, and people who care are people who have an idea what the paper is about.
It is completely wrong, and contrary to the principles of openness which make websites like this valuable to require any sort of expertise barrier, other than a moderately low rep threshhold just for getting rid of spammers and charlatans.