I have seen some journals/publishers sending a submitted paper to up to 5 reviewers and some other journals just follow a single or at the most two reviewers.
"The editor selects usually two (but sometimes as many as four) independent reviewers who research or practise in the same area as the author and are subject specialists..."
I think that 2 reviewers, who work independently first & then allowed to contact each other next, will be adequate. Each one has to be on full alert & the later exchange serves in better final evaluation or judgment.
Your emphasis is on the ideal thing to be done. I think it should be three reviewers. In that case, if two of them accept an article then the majority decision can be carried out.