Where have all the smart people gone? I've read several dozen posts and not one has pointed out that the problem of promising a scalar delivery date before determining subscription level can't possibly optimize over a metric of on-time delivery.
Kickstarter projects should be providing an estimated delivery date as a function of subscription level, where x1 (or less) is the number most projects now promote, but you also have numbers for x3, x10, x30, and x100. Out my ass, I'd guess you could fit a curve to existing Kickstarter data that would add six weeks to the deadline for each multiple of 3 in oversubscription level.
The Pebble project actually hit x100, so a realistic ship date in my mind might be 4x6=24 weeks later than the originally promised early fall delivery date.
The positive influence of having substantially greater funds to deploy (Pebble hired more people than originally planned) is wiped out and then some by the hugely increased risk level. If Pebble manufactures 85,000 watches, ships most of them out immediately, then discovers that 20% of the devices fail in under three months due to faulty moisture control or creeping solder whiskers, they might as well just blow the hole thing up.
Who is going to show up with $2,000,000 to bail them out of a huge PR fiasco?
One could say that the Pebble originally promised is late. Or one could say that the originally promised Pebble will never ship because it no longer exists. I tend to take the second view. The Pebble that ships 6 months downstream of the delivery date promoted during the funding cycle is not the same device. The manufacturing standards are higher, the QA standards are higher, additional features have been added (higher level BT standard, additional waterproofing), and the development environment should be further along (though I haven't seen any tangible evidence of this as yet).
The whole problem here begins with the phrase "The Pebble". "The Pebble" people thought they were buying/endorsing ceased to exist as the subscription level climbed toward the first $1,000,000 (the x10 subscription level). Pebble went deep into the regime of "a Pebble" from a spectrum of possible Pebble delivery scenarios.
The Pebble promoted was supposed to be manufactured in the S.F. region. The Pebble delivered will have been manufactured off shore in China. Until the subscription level was determined, we were truthfully talking about a Pebble modulo volume and risk. There's not even any point in totting up on-time delivery statistics without confronting the central fiction of the Kickstarter model.
When I signed up mid-snowball I viewed it as a quantum superposition of two entrepreneurial stories: A) a relatively low volume run with mid grade QA, immature tools, and a small target market; B) a high volume run with high volume QA standards, somewhat mature tools, and a moderately large target market for app developers.
The story was acceptable to me, either way. One watch, two stories. Kickstarter is not a single story engagement, even if the convention holds that only one of these stories is mentioned during project promotion.
A person has to be in some profound eigenstate of stupid, uniformed, myopic, deluded, distracted, self-serving, or litigious to fail to figure this out.