Organizing objects or people into classes is a cognitive process of grouping various empirical elements deemed similar in some respect and setting them apart from all other elements deemed dissimilar. This process of "sectoring" serves a function of "making sense" of empirical reality i.e. organizing it into meaningful categories, reducing uncertainty and ambiguity, and facilitating quick processing of information, making predictions and decisions in both science and every day life situations. Thus, there are always elements of objectivity and subjectivity in the concept of a class.
Therefore, the concept of class is theoretically useful only if it is based on some empirical reality i.e. groups elements that have some objective elements of similarity or dissimilarity with non-members, and if it has predictive validity i.e. can be useful in telling how th elements will behave or how that behavior will be different from that of non-members.
The concept of class based on members' relation to the mans of production fails the second part of the aforementioned criterion - it is too broad to have any predictive validity. That is to say, knowing a person's relation to the means of production (which is the empirical criterion of class membership) is vastly insufficient to predict behavior, even narrowly defined (e.g. voting). Other predictors are needed. Hence the analytical usefulness of the concept is equal to the share of variance it can explain - in this case close to nil.
The genius of Marx using ownership of the means of production as class identifier was its empirical clarity at the moment and also its political appeal - but that identifier was not the cause of what created the working class. If anything it was an effect. The working class was created by rural feudalism (as Barrington Moore and Co. convincingly argue) - which is to say that the class of landless and skill-less masses has its origins in the feudal or more generally rural relations of production. These masses and their social-behavioral characteristics - especially rural solidarity that was critical for creating the sense of social cohesion that made collective action, unionism etc, possible - were merely transferred from the country side to the cities and funneled into the lowest rungs of industrial production: the property-less labor.
This is why despite its political and empirical expedience at the time, the concept of class defined the ownership of the means of production, lost is usefulness as industrialism transformed social relations and social organization.
Today, political activists may see some usefulness in lumping together people of different walks of life into "working class" - but that effort fails on both criteria of class making - empirical reality test - there is too much dissimilarity among the elements/members for anyone to seriously believe that they form a coherent class, and the predictive validity of such a concept is close to nil. It is a futile search of the past that never existed - memories of one's childhood, community, the Main Street and good old times in general.
Then there is another issue of social inequality - as observed by Bartlett and others. It is an undeniable empirical fact, all right, but what do we gain when we call it in any particular way reminiscent of the old battle cries of the past. If that reviving of old battle cries was performed by the industrial avant-garde transforming the society and using the old battles to gain popular following (as the old man explained in the _18th brumaire_ ) - that would make sense. But if it is performed by the rear-guard fulminating over its loss of any influence and status - this is nothing but useless and ridiculous kvetching.
If the left is going to offer any convincing explanation of the currently existing inequalities, it must provide one that is based on fact rather than wishful thinking, and one that is useful, which includes inter alia being convincing to the middle class and mental workers - whose support is essential for any political victory.
I have a feeling, however, that many upper class radicals embrace "workerism" for reasons similar to those why some members of British nobility embraced communism - to enhance its own status. The nobility was threatened in its status by the ascending bourgeoisie thatw as also the dominant economic force. Thus conspicuous displays of wealth became noveau riche and parvenu. Therefore, the upper classmen had come with other class markers, one that bourgeoisie would be unlikely to embrace. One such status is the embracement of the social freaks against which the middle class defines its status - workers, criminals, prostitutes and sundry derelicts. Since the middle class defined its status by setting themselves against these derelicts - the upper class would define its status by embracing them, especially that there was danger of any confusion of being actually one of them.
In the same way, the US upper class radicals, eager to differentiate themselves from their bourgeois roots - embrace "workerism," gutter populism, and freak-fetishism to underscore their status difference. I would not find surprising - this is the way social status is created, after all - except that these radicals claim to reject the notion of status and status-based social differences.
Wojtek