It is a great way to get data, which is why I mentioned "except for quick tests".
For actual growth, it's actually a pretty poor channel. A lot of time and energy will need to be spent optimizing ads, LP, CTA, etc. Time spent on funnel optimization can be better spent running tests elsewhere.
I have personally found it to be far easier to pop a few emails and reply over the course of the day. People who are genuinely interested about the product will link to it (take for example this fine lady: http://preppypaleo.blogspot.com.au/2013/03/paleo-banana-brea...). Those are the people we want to look for.
That way we can more effectively run tests and grow users more naturally, as opposed to having a HN-level spike and not have any follow up activity.
The current tests we're running/actively developing for are pro-accounts. So instead of spending time obsessing over ad quality, CTRs, optimizing landing pages, we can spend more time working on writing the software needed to test our assumptions.
Hope that helps