The emphasis "never touch a running system" in general can't be ignored, point. But in this context, in polite words, this is a completely weak way of thinking. First we have to ask: Does it work? If so, which parts "work" better than others? From my point of view it all comes down to two points that make DNN the best and, as well, the worst CMS/WebFramework on the market.
First, what's great about DNN? It's the nearly complete framework that does support many scenarios with built-in APIs and and (core) GUI functionality. The high-level feature set is far superior compared to Umbraco or Orchard. There is simply less time to invest to get a client's site up and running. The outstanding ecosystem reflects that in a very useful way.
But what does DNN bad? Well, from my point of view, it's primarily the poor performance when it comes to more complex sites. Editing content can be like hell, even when you are a naturally patient person. The site visitor suffers from that as well. It's hard to keep the page load time under one second when a couple of modules are installed and active on the page - even on a well performing server. This can be slightly reduced by taking advantage of the new (but very basic) MVC support and a client centric way of development. Why this mature technology took so long to get into DNN is another question to be answered. So, here we are. There is a mature, proven and beautiful framework that lacks an important SEO and user experience feature: performance.
I think, in this context, a major design change and related (partial) recode and can improve the situation heavily and put the platform (back) on track for the (present) future. To be honest, I think it's needed more than anything else. MVC is the past, the present and the future, ASP.NET Core is the future for .NET based web development. My personal problem is that we, as a community, hear nothing about it. DNN neXt is off the roadmap, no communication since nearly one year. My trust is fading that DNN will push itself onto a new level to make itself the outstanding number one on the market. And this is a realistic goal, in my opinion. Just move forward, maybe in small steps. But move. At least communicate. Anything is better than a dead forum/thread. My suggestion is: Put it on the table again. Anytime soon!
Take a look on other projects that did a complete rewrite and those who didn't.