The two best-known features of Chuck Moore's style -- extreme simplicity and extreme factoring -- must be how he avoids the dangers you describe. I wonder if these qualities go beyond stylistic choice and are actually necessary if one is to build real systems in Forth. That might explain why programmers who come from other languages tend not to stick with it. Your story is reminiscent of others I've read in the past (most prominently http://www.yosefk.com/blog/my-history-with-forth-stack-machi...) that talk about being in love with the Forth model and then abandoning it because it got too complicated to make it do what they want.
Makes me wonder if trying to do what you want in that model is a recipe for disaster; instead, maybe you have to let the model tell you what you can do: if it's simple you get to do it and if it's not you don't. Chuck would see that as an advantage, but most programmers wouldn't.
Edit: the article I linked to talks about that at length. His conclusion is that if you're going to build systems in Forth, you need the freedom to shrink or morph the problem into something that can be solved simply in Forth. If you don't have that freedom (e.g. because people are feeding you inflexible requirements), it probably won't work...
Makes me wonder if trying to do what you want in that model is a recipe for disaster; instead, maybe you have to let the model tell you what you can do: if it's simple you get to do it and if it's not you don't. Chuck would see that as an advantage, but most programmers wouldn't.
Edit: the article I linked to talks about that at length. His conclusion is that if you're going to build systems in Forth, you need the freedom to shrink or morph the problem into something that can be solved simply in Forth. If you don't have that freedom (e.g. because people are feeding you inflexible requirements), it probably won't work...