A credentialed reader pushes back:
I'm someone who spends every day confronting the cost-benefit argument for the Olympics. The $17bn figure from the public accounts committee includes a lot of folded-in costs. For instance, the extra troops assigned for security were still going to have to be paid; it's just that now that's a line-item for Olympic cost because of how they've been assigned. It's a big number, sure. But a more accurate analysis would look at marginal costs, not just bottom-line costs.
Also, the economic impact shouldn't be measured in the short-term stimulus.
Preparations for the Games meant 40,000 people were hired or kept in work, most of them in construction, which is always particularly hard-hit during economic downturns. The regeneration of East London includes turning the Athletes' Village into 11,000 new homes – 3,500 of them meeting affordable housing criteria. The International Broadcast Centre will become retail and commercial space after the Games. Westfield Shopping Center is one of the largest malls in Europe. Improvements to infrastructure – especially transport and high-speed internet – are attracting new business to East London (see Tech City for the primary example).
There's no doubt the Games are expensive. But the headline number doesn't tell an accurate story, and the long-term benefits are nearly impossible to put a monetary value on right now.