When someone claims expertise in search engine optimization, and states categorically that something is an acknowledged fact by other experts, without any reference to statistical testing you are left with a few options.
Just believe them and stop doing what you have been doing up until now and subscribe to their competing philosophies of kumbaya blogging.
Carry on doing what umpteen experts have told you works for them, has worked for them for years, and most likely will continue to work until some monumental change in Google's algorithm
Go on the offensive, call it out as SEO bullshit, and see if they can come up with proof
Create a public test case to prove them wrong
Option 1 is for sheep
Option 2 has merit - just go about your business and feel happy that they are fleecing 1000s of people with their brand of cool-aid and enhancing your competitive advantage.
Option 3 is fun, maybe even good for short term linkbait, but doesn't really squash the other person's "opinion" without factual evidence.
Option 4 is the smart option
So we are today testing anchor text and whether it is passed as a ranking factor from an article directory to a destination page.
This isn't a statistical test, and in many ways is a little flawed
The anchor text we are going to use is unique to the internet - if phrases are unique to the internet, Google might be more inclined to count the anchor text.
The article might end up syndicated to other sites so it doesn't prove that the link from the article directory counted. However as the argument resolved around it being pointless to use article marketing for SEO, if an article is further syndicated then that is a benefit of article marketing.
This is just one test that could easily be compromised by a 3rd party whereas it would be normal.
I am not going to use article spinning to try to make each article submission unique, and I am not going to isolate this to just a single article directory as it wouldn't be fair to subject them to such scrutiny in public... I am sure there are some article directories which don't pass ranking factors from their pages due to paid links.
This is also subject to the whims of article directory moderators. It is a clear attempt to highlight part of their benefit that is frequently underplayed. "Pure" article marketing authors aren't writing articles just for SEO purposes. They love writing, it is their passion, they earn a living from writing no SEO, or write for widespread exposure of their ideas or talent.
There are also many article directories which use the microformat rel="nofollow" within their anchor links. This is even something I have recommended if article directory owners don't trust their authors, or don't have time to continually monitor existing links for changes in reputation. If your site links to 100K+ other websites, monitoring them on a continual basis for malware isn't a trivial task, even though there are technological solutions.