Have to side with trade-ya1 in this debate. (I've been in both buy and sell side of the hedge fund industry since '98). $100m is off the radar screen for most institutional investors. Performance attributes at that level are rarely replicated even at, say, a 5x level of FUM. Typically, if you ask a manager who's been working with $100m for some time what his estimated capacity is, his reply will include a caveat something like, "well, I'd have to adapt the strategies ..."
I'm surprised, Buzzy, that you consider running $6bn isn't much more onerous than $500m. Not only are market-related factors an issue, the organisation management burdens tend to increase exponentially.
If you can return 20% p.a. consistently, you will have people queuing up to place money with you. Marketing will hardly be necessary - your returns will get people knocking.
One might argue that slick marketing has been a huge factor in Man's success, but then they've also had attractive numbers to include in the collateral. When I was a broker/portfolio manager, it was admittedly easy to garner private investors on the back of some tortured simulated track record. As a marketer to institutions, you begin with having to counter a very visible cynicism the moment you open your Powerpoint....
I love healthy Sharpe ratios as a marketer, but they carry little weight with my due diligence officer. They are not difficult to massage. I believe that most people involved in researching and selecting hedge fund managers will tell you that the quant part of the d/d process doesn't weight more than 20% in their report.