Pricing for a listing is a very complicated process. The underwriters, along with the company executives, go out on a very long roadshow and book orders well before the listing date. It is through these orders, not some magic made up numbers, that the final price is determined from.
In the case of LinkedIn, they changed their list price no less than 4 times between the time they filed their S1 with the SEC and the time they finally listed. Initially it was $31, and the last hike to $45 only happen the night before the listing.
LinkedIn management are professionals who understand their business and understand the process, they have a choice of underwriters to work with and a choice of investors to take orders from. This isn't a single bank taking them for a ride - to suggest that is offensive to those who run LinkedIn
Where the real problem lies, and a problem that was not raised in this criticism, is in how orders are taken. This is what Google attempted to solve with their Dutch Auction system. The claim is that the banking community is so tight-knit that they collude with each other to keep the book price down. So what Google did was to hold a silent auction on bids and allotments, only to find when the process was over that most banks essentially bid around the same mark anyway.
Paying out 6.5% of your company to go public does suck - but it is the cost of creating a viable and flowing public market for your stock. You can't just sell that part of the company to 4 or 5 banks and then ask them politely to pass it on - you may as well just raise another private round in that case. The point of the IPO is to diversity ownership as broadly as possibly and to engage firms that would be willing to take on and trade the stock so that a market is created.
None of the alternate mechanisms work, and you need to be a very large and hot company to even challenge the status quo in the way Google did (and in a way LinkedIn did as well - with their two classes of stock). Note that the underwriters are taking a risk since they end up holding a lot of stock, and in the event of the list price dropping there would be a lot of questions asked about the prospectus and roadshow and potential lawsuits. Also on the other hand, there are not a lot of IPO's that take place, so the underwriters need to make the most of the business they do get - they are the ones with the connections to the large funds that purchase stock, so acting as a risk-bearing agent in that capacity does deserve compensation.
LinkedIn didn't help their cause by listing so few shares. When there is scarcity in the market and so much demand, then there is only one way that the list price would go - and that is up. If they listed twice the number of shares to meet demand then there definitely would not have been so much volatility on the opening day (a lot of which was caused by 'market' orders - which means 'buy at any price').
tl;dr: creating a free flowing and liquid public market for your stock is a very complicated, highly regulated and risky process and nobody has figured out a better way to do it