You can be very nimble with Enterprise Java. Plus you get the benefit of the biggest library bar none. The issue isn't the technology it's the mindset.

Look at technologies like Dropwizard or Vert.x.

Sure you can, especially with eg Groovy and Grails, but I wouldn't say it's Java's "sweet spot". Plus you tend to attract a lot of corporate enterprise Java devs who on average have a different mindset to quick-moving product devs.

I moved into a Perl shop straight after and the difference in productivity using off-the-shelf language features was staggering. Ironically I'm very interested in Clojure right now, I get the feeling it could deliver impressive productivity while leveraging underlying Java technologies.

There is now enough tech information (videos, stories on scaling & stack etc etc) freely available online that there should be no excuse for getting the tech foundations wrong. Unless of course the head of the fish isn't tech savvy running a tech company then you have a classic technology issue I have seen too often.

Bingo, the information was out there but the execs running the show didn't listen. A mentor with gravitas and a track record would likely have talked sense into them. This was 8 years ago though, so many more stories are accessible these days.

Unless that mentor was on the board it would still be hard to change a fixed mindset. But then again times have changed in the last 8 years.

Influencing people is a hard game. You can only do so much and the rest is inception. That's why you want the right people with the right skill-set driving the ship to avoid it grounding.