I presume she means the subscription model. Which, incidentally, is not just used by commercial publishers but also by not-for-profit ones.
I agree with her. But it's not the use of the subscription model by commercial publishers that is the 'primary problem'. It is the fact that the subscription system cannot cope with the unrelenting growth of scientific articles that is being produced worldwide.
Before the internet, the subscription model had increasing problems, but it was probably the least worst solution, by no means ideal. Now, with the internet working and pretty mature, we can have better systems. There definitely are publishers, for-profit as well as not-for-profit (just look at the recent press release of the DC Principles Coalition), who seem to be wedded to the subscription model, but not only publishers. Libraries, too, do not seem to be too keen on replacing the dysfunctional system with a better one. And even a school of thought in the OA advocate camp, the self-archiving champions, argue that the subscription system will continue to sustain journals.
Of course there are difficulties to overcome if one wants to make the transition from one system to the next, and let's concentrate on overcoming those difficulties.
The subscription system has the following problems (and quite possibly more):
The price to readers/libraries bears no relation to quality. This needs no further explanation, I guess.Most of these problems are solved in a system in which the 'publish or perish' culture (which is definitely not of the publishers' making) is reflected more transparently. A system in which research articles are seen for what they are: a kind of 'advertisement' in which the author 'advertises' his scientific prowess, in order to get acknowledgment, citations, leading to tenure, future funding, for a few the Nobel Prize, et cetera. That doesn't mean that articles aren't full of information useful to readers. But so are conventional advertisements.
The price to readers/libraries bears no relation to the amount per article that's taken out of the academic market. A 'cheap' journal can, on a per-article basis, take more money out of Academia than an 'expensive' journal. This is more common than is perhaps realised. A substantial number of not-for-profits have seemingly low subscription prices, but take more money per article out of the academic market than even the most expensive commercial publishers (where it hovers in the $5000 range). I know of several cases where it is twice or even three times as much, and if someone would care to analyse this information (it often is available, for not-for-profits), one might find even higher multiples.
The price to readers/libraries bears no relation to the cost of publishing, but rather, to the numbers of subscribers. This is the origin of the price spiral. Journals were cancelled, and for some reason commercial journals suffered more than not-for-profit journals, on the whole (with exceptions), as a result of which subscription prices went up. This caused further cancellations and thus the vicious cycle was created. One of the reasons why some not-for-profits have been able to maintain lower prices is the existence of cancellation-resistant compulsory member subscriptions.
The cost to libraries of subscriptions that are needed bears little relation to the size of the actual research or teaching efforts at the institute in question, but instead, reflects the width of the range of disciplines researched or taught. A specialised institute (take CERN as an example) needs no more than a handful of journals. On the other hand, a university where the name 'university' still relates to 'universal' knowledge, and where a wide range of subjects are taught and researched, needs vastly larger numbers of journals to satisfy the needs of its constituents.
Subscription price stability can only exist in an environment of stability of the number of subscriptions, and of articles published. But that environment doesn't exist. Library budgets have been under pressure for the longest time, which is especially apparent if they are expressed as a percentage of the research budgets. And the number of articles keep on growing.
The advertising analogy is not perfect, but I'm using it to illustrate the point that there is logic in the system that levies charges for the processing and formal publication of research articles and subsequently makes them universally available with open access. Open access publishing.