The base product was originally great: very smooth wysiwyg collaborative document editor with wiki-like linking. The problem is you don't need to do much on top of that. But clearly investors demand some "results" so PMs need to keep coming up with features that can be shipped in a quarter. Meanwhile bugs in the basic UX are plentiful.
Any really interesting work to improve the basic "collaborative document" experience is going to take time and experimentation, and I'm sure there's something to be found there. But the investor fueled focus on constantly doing something new and shiny means these really interesting spaces will never be explore and the product will continue to degrade with bloat each quarter.