Finding out what needs doing - through meetings, documentation and so on - is very much a part of software development unless there is only one developer and he is also the only stake holder.
This isn't even necessarily inefficient, because productively doing work that isn't needed isn't helping anyone either.
We had this "important project" coming up, a certain big feature that was wanted in the platform. A project which would take about 6-9 months with the collaboration of various teams across the company.
I was involved but only in a helping role. One member of my team was involved but they were fairly inexperienced and so I was asked to go in as support.
After the first two meetings a number of things became clear. Mainly that the product owner for the project was extremely ambitious, clearly not knowledgeable about the technology involved and got very upset about any concerns or problems anyone would point out. They wanted "those stupid meetings" to be just for informing them of solutions and progress, not for discussing problems and options.
At the third meeting I made the mistake of mentioning that a particular detail, not too important, would be extremely costly to do as asked and that we would be better off doing it in a much simpler form. Unbeknown to me, that day the product owner went behind my back and asked for me to be immediately removed from the project. So I was. I didn't care much for it and I only learnt about that move much later, when I had already left the company.
Anyway, the project ended up taking about 2 to 2 and 1/2 years and it had to really reduce its scope in the end. And some of the people involved confessed that it was because of the product owner, because they had wanted people to "just build this thing I say and stop arguing". A lot of work was simply thrown away several times simply because of not having had the proper discussions it needed.
This isn't even necessarily inefficient, because productively doing work that isn't needed isn't helping anyone either.