This is something I attempted to address some time ago, without much success.
I truly believe there is massive potential with this technology however the fundamental problem is that the average person doesn’t care about having complete anonymity. They want to send the odd email, do a bit of shopping, read the news and go on Facebook. To the average user it’s more of a risk going into an environment where everything is anonymous than not entering it. E-commerce (as we currently know it) is impossible - would you transfer funds to some completely anonymous entity? What benefit does the average person in the western world gain by having complete anonymity on the Internet? The average person doesn’t think/consider that if they don’t move into the SAFE Network they are sleepwalking into government oppression.
A few years back MaidSafe had a project called Sigmoid(x) (I think) which I though was a perfect use for this technology - but could have been taken further. If memory serves it was a project aimed at businesses so that rather than having dedicated file servers and a backup policy/system all data would be spread throughout the network, on end-user machines. I thought this was fantastic and something that gave real benefits. I’m sure all networks have very lopsided resource use and this went a good way towards remedying this, i.e. not having a few machines in a network being hammered (and having to worry about them going down) and all other machines sitting doing bugger all. Also since the capacity within the entire network is greater than a comparatively small number of servers is ever going to be it solves loads of problems - not least the difficulty in retrieving old backup data - typically if you’ve old data that’s moved onto tape then you pretty much should consider it lost, the cost of retrieval is rarely worth it.
I personally think the anonymity aspect is what’s being touted as the main selling point of this technology. I may be wrong but I think this is a mistake as it’s appealing to only a very niche market. Everyone wants to store/backup data, a lot of people want extra computing power, few people want anonymity.
The USP of FreeNet was anonymity and it (arguably) failed. Fair enough it is very slow and the SAFE Network may be much faster but I think if there was sufficient appetite for FreeNet then the performance would have been addressed.
In my opinion this network should focus on P2P (not anonymity). Cheap, fast, reliable data storage (which is hard considering how cheap this already is!) plus cheap, fast, reliable distributed execution (this isn’t super cheap yet).
Systems like HTCondor (University of Wisconsin) could be developed on the SAFE Network. HTCondor is successful but I think you could get something much more sophisticated and over much larger number of nodes.
System like SETI@Home are fairly successful however if you’ve got something like SAFECoin in the mix then you’re much more likely to have reliable nodes, making context switches (etc) less likely and therefore having better overall throughput.