High-powered incentives for the UAW?

WSJ reports that General Motors wants its unionized workers to have part of their pay tied to the company's financial performance.  From a financial standpoint, this appears to make a lot of sense on the surface, as worker pay would go down when the company has bad years thus reducing the variance of profits. 

I see at least two reasons why this may not end well.  First, GM stockholders and bondholders have multiple ways of hedging any risk associated with the assets they own, as anyone who has taken an MBA level finance course should easily understand.  Are profit-sharing or bonus plans desirable for UAW members?  These are individuals who for the most part have limited financial resources beyond their wages, benefits and home equity.  By putting more pay at risk, it is not clear what they can do to offset that risk.  Perhaps UAW members would have more stable jobs if the price of labor fell during downturns, but if I were a UAW member familiar with the long run trend of employment in the domestic auto industry I would be skeptical.  On the surface this looks like a classic case of inefficient shifting of risk from stockholders to workers. 

Second, how much would a pay-for-performance plan affect employee behavior?  This looks like a classic case of what I call the 1/n problem where individual workers see that the impact of their own behavior on the bottom line is virtually zero in any organization where the number of employees (n) is 100 or more.  Given the decades of mistrust between GM and the UAW, it is hard for me to see any new pay scheme changing productivity, especially in an industry where workers have little control over the pace of work and the connection between individual contributions and company outcomes is tenuous.