All were top balanced, with putting in parallel for 24 hours. They all settled to a resting voltage about 3.40 or so.
This is not top balance at all. This is very rough approximation of being at the same SOC, assuming all cells are the same capacity, which they are not in real life. Most likely his 5 cells were few AH less than the rest, perhaps being from a different batch. By just paralleling cells and not charging them to the upper knee together he did not accomplish much at all. Then, when cells of slightly different capacity, but same SOC were put in series and put to use, smaller ones drained before others and took some damage. From that point you are on the down slope as damaged cells keep accumulating damage since you can't keep them in line with the rest since you have no tools attached to them.
Tracing back this story one can assume that 5 cells were simply few AH below others, but as long as they were not less than rated capacity, you can't claim they were defective.
Simple cell level monitoring tool would have clamped usage of smaller cells at their LVC level and prevented them from taking a dive. Alternatively, if he bottom balanced the pack to begin with, they would all drained together and not gotten damaged, but this means 5 smaller cells would shoot up at the end of charge and would require charger adjustments to make sure it stops when those 5 are full. All of this also assumes he knows how to bottom balance correctly, and seeing if one can't top balance correctly one may not be any better at bottom balancing.
The issue here is that pack management terminology is discussed until everyone is blue in the face, but when it comes to practice many tend to do it wrong and then blame the cells.
Spdas, please do not take it personally. I made bunch of assumptions here based on what you said, which may not be the actual story. I do not discount the possibility of defective cells, its certainly possible, but when you refuse to use simple tools the you can't prove anything one way or the other.