Perfectly right. Never take the word of anyone on Internet, always use your own criteria, is the best that you can do. What I suggest to anyone, if you find something on the internet against your bias or knowledge, try to find why it is different, and in which circumstances it may be true.

In my case, I only have 35 years on developing software, in really different markets, but because I share it with the task of being professor in the university, I have the opportunity to gain some perspective.

I agree with many of the things that you say, but not everything. It is good to have it here, so others can enjoy of two points of view, try to understand each other, and grow. In my case, I have already been in some of the stages that you comment, and if you could look for old opinions, you would see that I couldn't agree more. But in the past. For example, in less than 10 years I changed to be a total detractor of testing, to not be able to live without it (with some clear exceptions).

As an example: the TTM that you comment says that perfection has a cost, 100% agree. In fact, my talks about automatic-testing for VPs starts with that fact. But I do not know if you are implying that automatic testing is costly because testing pursues the perfection. Because if it is the case, testing does not guarantee perfection. If you want perfection, you should use the mathematical demonstrations created by Dijkstra in the 60'. Those demonstrations do not work because the most important is having the right feature, not a random feature working perfectly.

And just because I think that it is good to know, a company like Tesla that creates physical goods, can release a new idea in just three hours thanks to TDD. That is a superb TTM.