I wouldn't say that there are any clear advantages. In fact, i'd go for whatever form of advertising comes out cheaper, yet accomplishes your goal.

If the cost of keyword advertising is more than display advertising (barring some other considerations like lifetime value of the name from the particular channel you advertised) you should probably go for what's cheapest (at first).

In the begining it's all about testing different avenues to find what works. The more channels and lists you can test at once, the faster you'll get to profitable results.

Let me add one more thing, from a business standpoint I feel that keyword advertising works better. But this is if you're looking at your business from a publishers standpoint.

In the publishing business, your goal is to collect e-mail addresses as cheaply as possible so that you can turn around and sell them something they might be interested in. You look to make up the money you spent by selling them products.

Over the course of time, you track where your names came from and then start looking at which names provide the best lifetime value. Once you figure out the most profitable lists, you can continue going to them over and over.

In the business we are in, we still haven't gotten display advertising to give us the same results as keyword and e-mail advertising. Well, we've had very sporadic success. But nothing overwhelming.

We've also found that test buys for company's like Yahoo have very expensive upfront costs. Where as in Google, we can spend $10,000 here, $20,000 there, etc...

In the end, marketing - whether its display ad's, keywords, or e-mail ad's - is all about testing and seeing what works.