Some states, like Texas, are extremely picky about who calls themselves an engineer, irrespective of industrial exemption (which the vast majority of people fall under). PE licensure in other engineering areas (like civil engineering) is required to offer services to the public, helps when offering expert testimony, and can be required for principals at consulting firms. The software engineering PE is an extension of IEEE's efforts in software engineering, and is probably a good thing for safety-critical industries.
Since passing the fundamentals of engineering (FE) exam is (usually) a prerequisite, someone who has a PE in software engineering probably knows their way around a free body diagram or Laplace transforms or MSD systems, i.e., stuff the other engineering fields do.
So, if my FE is still valid, I could look into the PE for software? I'm intrigued.
EDIT:
Link for Texans: http://www.tbpe.state.tx.us/software.html
While I take the point that there is a difference between code-monkeys who don't have deep understanding and serious well grounded engineers, I don't agree that a two year old exam in a field that's had a history of failed certification attempts is a reasonable benchmark.
I think part of the problem in the field that contributes to the article is that the field is young and we don't have strong, well accepted professional organizations and licensure yet. We're still figuring out what those look like, and dealing with some serious challenges in breadth as we do so.
The "software engineering" PE is apparently pretty new (first offered in 2013) and kind of an oddball since this industry didn't grow out of any engineering tradition and doesn't really place too much value on certification.