There is no doubt that double opt-in, i.e. where a subscriber, however they subscribed, is sent an email asking for confirmation that they wish to receive your offers, gives better results that the single opt-in. All research confirms this, the only dispute being by how much.
However, most samples are taken from a range of businesses and, as with everything in email marketing, there are variations. You have to discover what is best for you.
If they had limited the sample companies by, for instance, size, it is probable that the figures would have been somewhat different. For a small business just starting an email marketing list, a single opt-in might have benefits.
For those new to email marketing, there are distinct advantages to building a list quickly. Until you have enough subscribers to provide a statistically significant total, you will not be able to depend on your returns. One of the criticisms, and a valid one, of having a single opt-in is that the metrics will probably be slanted, but on the other hand, a wide range of interpretations is better than none at all.
If you accept that the results of such a campaign cannot be completely trusted then at least you have something to work on.
Another valid point is that hard bounces can be a problem. Many, although accept that by no means all, can be eradicated by sensible systems when collecting subscribers. Unsubscribe rates are higher with single opt in but if you have very few subscribers then any rate is worrying.
Whilst it is certain the double opt-in is probably the best model for established businesses, it is probably not true for all. Most importantly, no one knows what is better for you. You will have to experiment for yourself.
One way you can get the best of both worlds is if you include the option of a double opt-in on your first email marketing campaign to a new subscriber. The returns for those who fail to click on the link can be compared those who do.
There you are, your first dependable metrics and a stepping stone to many more.