Before there can be redistribution, there must be distribution, and if the initial distribution is unfair - and it's hard to argue that, for example, financial executives were paid their marginal products over the last decade instead of being paid inflated, bubble based incomes - then there's no reason to suppose that redistribution necessarily makes us worse off.
If we accept that being paid the value of your marginal product is fair - and not everyone would agree that it is - but if we do take this as our standard, then we need competitive markets to assure that wages are held to this level. Without the discipline of competition to regulate the marketplace, profits, and hence the income of those at the top of the income distribution, will be larger than can be justified by the contribution those individuals make to output. If that is the case, if markets are not ideal and there are significant departures from pure competition, then redistribution can improve the outcome in terms of providing the correct incentives in the marketplace. Redistribution is not necessarily harmful (that is why, for example, modern macromodels with monopolistically competitive agents often assume that the government imposes a set of lump-sum taxes and transfers to correct any distortions arising from the presence of pricing power).
The point is that those who argue against redistribution are implicitly assuming that markets are competitive and that people are paid accordingly. Thus, any interference is a distortion. But if that is not how a typical market functions, and I'm not convinced earnings at the top end of the income distribution are set in anything approaching a perfectly competitive marketplace, then interference can take away distortions rather than create them.
If people want to make the argument that incomes at the top are, in fact, equal to the contributions these individuals make to output at the margin, then they should make this argument explicit. Tell us why you believe this is true. But I suspect that a lot of the people who would be tempted to do so are the same individuals who were, not too long ago, arguing that financial executives were paid according to their contributions at the margin. They may still want to make that argument, but I think it's pretty clear the incomes these individuals earned were excessive, and that redistributing part of it would not have created bad incentives. In fact, it's easy to see how redistributing the excessive gains might have helped to temper the recent mania and improved the economic outcome.