Hello!
I've started experimenting with SDD quite a lot in the past months from easier to complex tasks, to try to utilize agentic coding in a way that results in a more efficient dev workflow. I see a lot of benefits a part from efficiency, such as better documented features and higher quality outputs, mainly due to spending more time planning and analysing the tasks.
At the same time, are there any studies showing empirical results of this workflow, or is it too early to ask? Would be interesting to see what the average net-effect is, especially due to the increase cost of AI-usage.
I'm strongly believing in this new workflow from my own experience, but I can also see the somewhat new bottlenecks popping up, especially during review processes etc. Though, it'd be interesting to see some hard facts.