If you look at operational costs per se, and licensing fees in a scenario like that, you're missing a huge proportion of the most threatening costs. Consider hospitals, for example, and recent problems with ransoming electronic health system data.
What concerns me with SaaS is that the biggest costs are very catastrophic or serious risks (like lock-in dependencies, or hacking due to monocultures--by the way, is that a formal, modeled concept in computer security, like it is in agriculture? The vulnerability of a market to "infection" due to monocultures?) that are maybe high probability over a very long-term period, but very low over a short-term period, and hidden.
As the competition heats up and providers become more efficient, I actually expect prices to go down even further.
However, as a software developer, calling it rent-extraction glosses over architectural differences and value derived from not having servers tucked away in an office closet. Flexibility and capacity on the server-end, as well as huge benefits of software distribution over web-browser reducing support costs. I derive a ton of value in never having to drive to a data-center and paying EC2 by hours-used seems reasonable since there is actually a server (albeit virtual) running for my money.
However, it's certainly rent-extraction when the only thing thats going on via the Internet is a licensing check.
Adobe Photoshop switched to the cloud model, so graphic designers now pay 30-$70/month for the entirety of their career. Yeah it's always an up to date version, but before the shift to a SaaS model, it was some $600 but it would keep working for years if you couldn't scrape together the cash for the latest version.
SaaS is indicative of greater concentration and the possibility of oligopoly, though. I think it's only a matter of time before the cloud platforms begin coordinating price rises beyond the value they provide.
If they try to regulate on-premise software then it's time to break the glass.
Look at Oracle.
You could move all new correspondence to a new (eg: on premises) system - but you'd have to keep paying to retain the same access to your business history.
Take any of the many business/crm-apps, where there's no clear migration path (you'd want both history and business process rules) - and it's easy to see how SaaS might easily facilitate "rent extraction"?
As others have mentioned the problem isn't unique to SaaS - but with an on premise solution, you might be able to move some terabytes of data by simply moving some physical drives - or at least have local access via you SAN. With a cloud provider, many businesses would struggle to get at the data at a higher rate than 100mbps even with the help of the SaaS provider.
And there's often no option to keep access to "old" system/data without paying rent, unlike some on premise solution (keep the old system, on it's own network segment, ticking along for historical records (or: in a VM).
And that's not a problem specific to SAAS.
Good luck getting 1-10 Gpbs data transfer from most public SaaS to your own datacenter without having to pay over and above your current fee for Internet access (eg: have a look at what it costs to get just a TB data out of S3 -- on any budget (or prime) dedicated provider, 10-50 TB of data would be included - per month.