I'm calling it now that FD drops P2P.
The more I think about it, the more I tend to agree with you.
Yes, they could make P2P work well enough, but the solutioning costs & timescales, ongoing dev costs of defeating exploits on local clients and the fact that there will have to be a such a large amount of centralised security & integrity checking anyway, coupled with the massive reduction (up to 90% in some cases) in cost of deploying dedicated game servers on elastic compute instances (Azure or Amazon EC2), will eliminate any possible financial gains that may have been made in the long term by going P2P.
The opportunities for this sort of hosting provided by cloud computing will only improve, accelerate and cheapen over time, the cost & effort-time of labour to code and maintain a complex P2P solution will remain roughly the same. Possibly FD thought of this idea when game companies were still deploying their infrastructure on physical kit, but those days are gone. No one deploying a new game deploys physical infra, in the same way no one deploying any new business solution would deploy on physical unless there were extenuating circumstances, usually related to data protection legislation across borders, or extreme security requirements.
This is why MS are offering free dedicated servers for all their game devs on XBOne, it just makes sense in 2014.
P2P is not a good strategic decision IMO and I think that beta may well prove this point.
But that is what tests are for, proving or disproving theories and giving people an opportunity to remake decisions.