My question is two-fold:
1) If this is the age of iterative, and small releases, how do people pull off big releases? Are they in essence waiting too long to release the many small parts that make up the big release in order to get attention to the big release?
2) I sometimes hear startups say "we have a private beta with 1,000 users..." and I think to myself: "How did they manage to sign up 1k people when even in a very open environment, it is hard to get 1k sign-ups.
release early, release often.
http://en.wikipedia.org/wiki/Release_early,_release_often exactly as you said, iterative small releases. however LIMIT the number of users allowed.
release 1 - 10, release 2 - 100, etc...
then before your big LAUNCH do two things:
release early, release often is not the only strategy. It may be considered as an option in the early stages of a web startup.
There are many ISV who make regular releases: quarterly or yearly. This is due to their contracts, partners and most importantly due to the way their customers upgrade.
And then there are ISV who do not release until a large number of features are introduced. This is to make sure customers have enough incentive to purchase support and maintenance contracts.
As your business grows, you'll discover that feature planning and delivery strategy is far more important to your revenues and computer-science based strategies like "release often, release early" are way to simple and naive.
I think release early, release often is better for new startups. For a new company if you have too long a release schedule clients are less likely to be attached to your software and go elsewhere, or may think you have done dark.