A study published in 2007 as part of the EC's push for one share share one vote found a rather complicated picture. There really is no way to sum up this paper in a paragraph because there are so many contradictory findings from the studies that it looks at. But I think it's reasonable to point out that the idea that things like differential voting rights inevitably lead to poor performance (because of entrenched boards pr whatever) is not easily proven.
There's also an interesting section in the middle which I'll post below, where an old quote from Adolph Berle, of Berle and Means fame, arguing that the rise of institutional investors (and therefore intermediation between the ultimate beneficiary and the name on the share register) had arguably already split voting rights from cash flow rights:
Now this stock certificate, carrying a right to receive certain distributions and to vote, begins to split. Once it is bought by a fiduciary institution, be it pension trust, mutual fund or insurance company, that institution becomes the “stockholder,” holds legal title to the stock certificate and to its right to vote. But it has by contract dedicated the dividends or other benefits to distribution among beneficiaries under the pension contract, the fund arrangement, or the insurance policy. The one remaining power by which the recipient of corporate profits might have direct relation to corporate ownership has been divided from the benefit itself.