Second of all, super-weeds is a very loaded term, and is a bit mixed up here. What happened is that the gene construct seems to have jumped from Brassica to charlock in a single plant, which wasn't able to produce seed. Contrary to what the article claims charlock is a relatively close relative, being in the same family Brassicaceae. You can even get that from the name - all mustards are Brassicas, and charlock is also called field mustard. The current Latin name for charlock is Sinapis arvensis, but it used to be Brassica arvensis!
What the article mixes up with using the term "superweeds" is that (practically identical with antibiotics in humans) you create new selection pressures for other plants by using herbicides. When you seed GM-plants you can use new herbicides (or use these more indiscriminately) which slowly creates weeds that are resistant, just like we made antibiotics-resistant bacteria in humans. These are now commonly called "superweeds" (again, a very unlucky term). You could leave GM out of the picture and say that these weeds have been created by herbicide (mis-)use.
Since 2005, many more cases of gene flow between GM plants and wild plants have been observed. Are these "bad"? In this case herbicide resistance is not very beneficial to the host plants outside of fields, so these plants are outperformed by others without the modification, as they're wasting resources. Is this bad for the species? Cross-species gene flow happens, and it depends on the case whether it's "good" or "bad".
Actual paper: http://cms.daegu.ac.kr/sgpark/plant%20biotechnology/GMHT%20t...