Epicycles aren't wrong, they're not even wrong. You could model an orbit that looks like homer Simpson's face with epicycles, but the ancients weren't sophisticated enough to understand that. The problem is that the technique of epicycles this has diminishing explanatory power, as you add more free variables (how many epicycles you are "allowed" to have.
The reason why LCDM gets called epicycles is that there is charitably one free parameter per galaxy (dm: bm ratio -- but actually more, because it's an arbitrary distribution)... So the fact that it has high explanatory power is unsurprising. And what's should be dismaying is the number of things it can't explain, given just how many free parameters you're allowed to have in a DM model.
The reason why LCDM gets called epicycles is that there is charitably one free parameter per galaxy (dm: bm ratio -- but actually more, because it's an arbitrary distribution)... So the fact that it has high explanatory power is unsurprising. And what's should be dismaying is the number of things it can't explain, given just how many free parameters you're allowed to have in a DM model.