Yes, I get all that. You're repeating stuff that has been talked to death on this forum. What I'm not getting are numbers/sources to what you said. Do you actually have any of that, or is what you posted just a feeling.
The data is available in the ICANN registry reports and the zonefiles. It takes a lot of expensive work to crunch that data and more to analyse it properly. I'll explain it in simple terms for you.
The velocity of net growth (new registrations - deleted registrations) in .COM had been slowing for a number of years and the monthly net growth had become rather cyclical with large promotions leading to large drops the following year. The Chinese bubble changed things and a lot of the new registrations appeared on Chinese nameservers that have low one year renewal rates. When the historical renewal rates on these hosters' nameservers are checked, there are clear patterns of high non-renewal suggesting discounting/promotions and short-term and speculative registrations.
The .COM has become a legacy TLD in many country level markets. In some the new registrations are just ahead of deletions. This is because the local ccTLDs have overtaken .COM as the dominant TLD in those markets. This means, in simple terms, that .COM is not doing as well as it was in those countries. The one country where .COM completely dominates the market is the US. Got that? Good! This means that the US is the most important country market for the TLD.
Any competition in that market is not a good thing for the .COM TLD. The .US ccTLD only has a minor share of the market and is so poorly marketed that it is not a major threat to .COM.
The new gTLDs have a novelty factor. And there are also some US city gTLDs that are getting registrations. At the moment, they are low numbers and they are, based on web usage patterns, (Yes. I have the web usage and development percentages for the new gTLDs.) developing slowly. Other new gTLDs are also getting US registrations. The development of these new gTLDs will be, for most, a slow process but they are part of a switch from the generic to the specific. The generic being the .COM extension and the specific being the niche extension of many of these new gTLDs. Taking these new gTLDs as specific, niche extensions in the ccTLD pattern rather than generic TLDs, means that they will develop as such. They have the capability to focus a lot of specific registrations from .COM. Initially the registrants will have their .COM (or .ccTLD) registration and will point the new gTLD registration to their existing site. But if any of these new gTLDs build serious momentum, that will change and the .COM (or .ccTLD) registrations of the registrants will be pointed to the new gTLD as the primary brand. This is what happens in ccTLD when a country level market moves from the early phase (where .COM dominates) to the ccTLD dominant phase (mature market). It can take years for a market to mature and all new gTLDs are in their early phase. It can take anything from five to twenty years for a market to mature. Some new gTLDs may never break 100K registrations. Others have, despite people thinking that they would not. The advent of the new gTLDs has made the market more complex and has introduced competition for .COM in its "home" market in a way it has not had previously. The .US ccTLD was so poorly marketed that it was never any threat. What has happened with the new gTLDs is that it isn't just a single TLD launch every few years as per BIZ/INFO/MOBI etc but rather competition in from a large number of niche TLDs.
That should be a simple enough explanation for you.
Regards...jmcc