Why did they disallow dotless domains? Honest qyestion.
As an alternative, one could write and distribute a "free" web browser with a search box that looks like an address bar where the search goes to a particular search engine that sells ads. One might call it "omnibox" or something like that.
Or one could even add code to the browser so that a dotless domain typed into the address bar automatically submits the string typed to a particular search engine and sends the user to the website that is the first search result.
It seems like these alternatives are not enough to capture all address bar traffic... because some companies that employ the above alteratives (you know who) still applied for new gTLDs that appear to be valuable only as dotless domains.
In other words, it seems like companies want to use all three strategies to capture address bar traffic:
1. A dotless domain typed in the address bar redirects to first search result of a particular search engine.
2. A dotless domain typed in a search box, that could be (and is) mistaken for an address bar, searches a particular search engine instead of causing a DNS lookup.
3. A dotless domain typed in the address bar of a browser that does not do 1. or 2. (e.g., an alternative to the major "modern" browsers) performs a DNS lookup where search engine/browser/ad sales company is the registrar for the dotless domain looked up. That is, they paid off ICANN ($185,000+) to have sole control of what some call a "public resource" (domain address space).
There are other strategies that can also capture strings typed into the address bar, such as the NXDOMAIN redirect nonsense, but I'll stop here.