In my opinion, ExWhyZee should have their registry pulled by ICANN for their unethical practices. They are using (without permission) everyone else's domain holding's to determine who to "give" the ExWhyZee domain to and then ABUSE that person(s) further by making them become the proxy registrant and having that name redirect to a parking company selected by the registry and for the sole benefit of the registry.
In other words, the registry is essentially "tasting" the domains free of charge...a practice that is now banned by ICANN.
The .XYZ thing is really just a land grab for registration volume and it is reflecting badly on new gTLDs as a whole. The logic is that at least some of the "registrants" will renew these robot registrations. It is a very high risk strategy and it generally results in a boom and bust registration/renewal cycle of the type seen with .INFO gTLD.
Taking the results of this survey of all the new gTLDs as a single set of TLD results is wrong as it allows the .XYZ anomaly to destroy any reasonable conclusions. The level of PPC in .XYZ is actually 90.27%. I ran a survey of the top ten new gTLDs on 26 June (and a set of 110K domain surveys of .COM, .CO, .EU, .CO.UK, .FR, .DE, .ES in early July) and it seems to have kicked off a bunch of me-too surveys. The Verisign results are reasonably accurate on a TLD by TLD basis but the categories are wrong because of the reliance on Eurid's "categories". Eurid's attempts at web usage surveys were as dodgy as its land rush. It only surveyed 5,000 domains and waffled about how good .EU was doing. It wasn't and the categories were cargo-cult used to make the ccTLD look a lot better than it is doing in reality. Again I ran a 110,000 domain survey of that ccTLD and the results were somewhat different from Eurid's results. Now it seems that ntldstats is trying to jump on the bandwagon and hasn't even properly defined categories. Measuring web usage and development is a very tricky thing to do right. It can often be very upsetting for registries to find out that all of their domains do not have active websites. The .CO ccTLD was a classic example - it had paid some company to provide a web usage survey and the results, methodology and categories of that "survey" were utter rubbish. It was even counting tens of thousands of redirects to Godaddy's PPC lander for undeveloped domain names as ordinary redirects. It did make the .CO look good until someone with a clue took a look.
To get a real view of web usage in the new gTLDs is necessary to consider each new gTLD independently of the others because they are, theoretically at least, targeting different markets and different users. What must be far more worrying for some of the new gTLD registries is how the daily registration volume has dropped below 100 domains for many new gTLDs. If these registration patterns are sustained, many of these new gTLDs will go the way of the small repurposed ccTLDs. And the registry landgrabs of premium domains are actually having a far more insidious effect on web usage and development in these new gTLDs.