As far as I know, there is no threshold to the number of dimensions, measures or marks that will cause a problem simply by the number of items and has a lot more to do with how the query is formed, how much data is being processed, computer specs and many other things. That being said, you mention 150+ column headers, are these all in the Viz or simply in the data source? If they are all on the Viz, I would expect that to cause some slowness but if they are just in the data source, I would not. I would suggest setting up a Performance Recording to better identify the root of the slowdown: Create a Performance Recording and Interpret a Performance Recording Workbook.
Hope that helps!