In 1998, I went to the first talk at OOPLSA about eXtreme Programming. It had been all the rave on the original W2 Wiki. I spent the next 5 years really getting into the movement, experimenting with it, and participating in the larger movement being labeled as Agile. One of my favorite books was Beck's original eXtreme Programming eXplained. It was pretty simple, pragmatic, and it didn point out that the practices kind of supported each other. I found it made a positive difference in the product I was working on at the time. Whether that was a placebo effect that really boiled down to "let the developers think they're doing a new cool thing and managing their own destiny and they'll actually increase their productivity" or not, is something I guess I'll never know.
What I have watched over the last 20 years has been like an accelerated variant of the last 2000 years of Christianity. The amount of pressure to be "doing agile" was huge, but people mixed and mashed it with their own ideas. Which had a variety of mixed results. The "introspection" principle (the idea that you should sit down at the end of every week or so and decide what was working and what wasn't and adjust as necessary). Like the term "Christian", when diluted to the point it has been, it becomes pretty undescriptive to use "agile" as a term to try and communicate what it is your team actually does do (or fails to do).