It is desirable that Questions posted on Math.SE be fairly self-contained and definitively answerable in terms of reasoned mathematical argument.
Naturally these goals are not always perfectly met, but the procedure you suggest -- posting an article elsewhere and asking here for a critique by providing a link to it -- seems at least open to abuse, even if your motives are good.
I think it may very well help you to learn from posting Questions and Answers on Math.SE to articulate and refine your mathematical thinking. That much is certainly consistent with our mission.
The distinction I'm drawing is principally one of granularity. Let's say you want to write an article on finding roots of polynomials. You may have in mind covering ground that is familiar to high school students and then following up with some more advanced topics.
Linking to a broad article and asking for a review puts a huge burden on Readers in comparison to the amount of excellent content that would get added to Math.SE. In fact a substantial critique of short length would probably require future Readers to consult your off-site article in order to get much comprehension of the points made. This borders on intending to drive web traffic to your site and away from Math.SE, and this is at best a conflict of motives.
On the other hand your article might reflect an arrangement of the typically narrower Questions that Math.SE does a good job with into a more comprehensive guide to a broad subject (like finding polynomial roots). In that spirit it might be (given appropriate links and authorship acknowledgement to Math.SE contributions) a project that benefits both you and this Community.
So I'm suggesting that you consider digesting material to suitable narrow posts on Math.SE, where the setup and problem to solve are reasonably self-contained, after which the results may be used with Creative Commons BY-SA 3.0 permissions in an article posted elsewhere.