@Kental2: This is why you need a dedicated project manager or account manager. Developers loathe them most of the time, but they are good at deflecting a lot of the vagueness. Developers brains are far too pragmatic and advanced to enjoy all the coffee and chit-chat that goes with the decision makers. I have come to accept that.
I have worked at a Russell Group University here in the UK reporting to a Professor and mentoring some PhD's to develop a forensic system, I have also worked at Agilent on some DNA, RNA and Protein software, the place was wall to wall with PhD's, coming to think of it I have also worked at another Life sciences Company but I digress, in each and every instance I have never had anything to complain about with regard to project management.
A scientist can spend 3 months on a project and discover that it just does not work, and have to start again. This leads to highly organised people and projects that manage themselves.
Alas, every company is now a software company, and you have a lot of people that could never get on an engineering course if they tried, getting involved. This really is where a lot of the problems are. I favour scientists because even on tough projects, they "factor in" failures, something the commercial world tends to treat as a disaster.
In the image below, the difficulty is in telling the customer that android was delivered in several iterations over many years. Customers find it nearly impossible to ever accept that a software product could be like the first version of android. A phone that just made calls. That's it.That's all it did. It was version 2-3 that they added a camera, and still then that didn't have a camcorder.
This is why developing software is hard for the most part, if you can manage your customers in a way that you have clearly set objectives, and Version 1, 2 then 3, then a lot of the headaches simply go away, as customers always want to launch with android kitkat or jellybean first.