There’s an article at Bridging the Gap, asking this question.
For me, the current set of best practices are indeed the best, at the current point in time. That’s not to say that they will always be the best, and it can be hard to spot when things should evolve – especially when it’s not you who has changed, but your clients or technology has evolved and your expectations and practices need to update themselves.
You can pick and choose from the various tools and methodologies for the best approach to the problem, and not everything may be appropriate, but what you’re choosing from should stand up to scrutiny and you can justify why you are or are not using the various techniques.
Picking the best bits is especially an issue with methodologies. Although ‘Agile’ might be what people are interested in, the entire Agile process might not be appropriate for every project. You must be able to pick the best and most useful bits for the client.
However, I can’t imagine a situation where I would voluntarily revert to known ‘not best practice’ approaches. Usually I’ve learnt what is considered best practice after many trial and error efforts. I’m an eternal optimist and I’d like to think that things only get better.