yes and no
for online growth you should build a system, not a website.
but a system does not necessarily mean multiple webproperties. running, promoting, maintaining a successful, growing webproperty is a full time job. if you start out with multiple webproperties you will do to much, to soon and you will fail. i see this again and again with publishing startups (here in europe) that target multiple markets or multipel languages. running two interconnected webproperties (i.e. language versions) is not double as hard as running one, but hard². a webproperty must definitely be seen as something that must scale and something that you can "branch out" from (diversify) but start with just one (1) thing (market, language, webproperty). if you found what works, then do more, do them differently and make sure that the parts make sense when seen as a whole.
ad risk distribution: yes, if you have just one strategy - i.e.: one webproperty and have an aggressive risk startegy you are betting on just one horse, and you are running it hard and at one point, it will fail. so yes, hedging is necessary.
in the past hedging (in regards of traffic growth) was meant having multiple webproperties, multiple domains. because search was basically the only way you could drive massive and growing traffic to it. this is no longer true: native apps, push notifications, social push, social shares, mail push (i.e. newsletters, email alerts) can now deliver enough (traffic/user) growth that your company can thrive and succeed. this is already hedging - and thank to zuckerberg and jobs and mailchimp (and other) we now can hedge our traffic/user (growth) strategies instead of needing to hedge our SEO strategie (from the start)
tumbs up for the Donella H. Meadows reference, she was the ultimate growth hacker decades before the term existed.
