While it is true that beer will naturally clear in a primary if given enough time (three months is a long time), most of the less flocculent strains will not clear as quickly as they do when racked to a secondary. That's why brewers have resorted to cold crashing.
Is a secondary fermentation vessel necessary? Well, I contend that brewers who serially bottom crop should seriously think about racking before the beer fully clears. The choice to rack to a secondary fermentation vessel or straight to a keg is their choice. However, bottom cropping well-sedimented beer carries over the least flocculent cells. The least flocculent cells are often, but not always petite mutants. That's a problem that the practice of cropping from a cold-crashed primary has compounded.
I also believe that a secondary fermentation vessel can be an advantage early on in the game when one is still bottling without the aid of cold storage. Under those conditions, a secondary fermentation vessel will get clearer beer one's bottle much quicker than waiting for it clear in a primary fermentation vessel at fermentation temperature. Let's face it. New brewers tend to stick with BRY 96 in one its forms (often US-05) because the strain is so forgiving. BRY 96 is not the most flocculent strain, and new brewers have not yet mastered the art of racking; hence, a lot of sediment gets carried over to the bottling bucket, less sediment in the fermentation vessel leads to less sediment in the bottling bucket.
By the way, the 3-month test that hmbrewing ran proves not only that autolysis is overblown, but also that oxidation when racking to a secondary fermentation vessel is overblown. The old school autolysis bogeyman has been replaced with the new school oxidation bogeyman.