Two schools of thought with estimating time frames:

a) under-estimate the time and then continually revise the date as delays happen. "hey, we would've had it done if x, y and z hadn't occured!" In this way, you create optimism and excitement (and maybe increased share price) and then drop the delay bomb later once people are locked in.

or,

b) over-estimate and deal with the backlash at the start. "there's no way in heck it should take that long, management is totally incompetant" Share price suffers, investors are pessimistic.

I guess you just have to pick the lesser of two evils. Clearly Liberty chose option (a). In the business I'm in, we choose (a) as well. Better to let the customer down once they're locked in than to lose the sale at the start.