The DM distribution itself is effectively a huge number of parameters. You have to have just the right amount of dark matter distributed differently in each case to explain observations and get those best fits.
Yes, the cuspy halo problem wouldn't be a problem at all if you were to simply adjust the infinite parameters that you suggest that DM has. The fact that there are statistical discrepancies proves my point that no one is adjusting a huge number of parameters.
And they still fail to explain other observations which require additional arguments as to why the dark matter is once again distributed in just the right way to give the rotation curve... that is already successfully predicted by MOND in most cases with just one universal parameter.