I'm the founder of Complice, one of the sites featured. It took me a bit under 2 years to reach my "poverty threshold" target, working an estimate of 20h/week on average. To break it down further, it was more like 5-15 hours/week for the first 7 months, then 30-50h/week that summer when I was off from school, and other similar fluctuations.
These are all total ballpark estimates that I made a couple weeks ago: I've used various time-tracking systems but none that give accurate overall time spent. Based on my within-Complice stats though, I can tell you that I've done 2800 tasks towards Complice. That's not very specific, but still kind of a neat figure.
I don't have a good breakdown between programming and marketing time, but I would estimate that 80-90% of my time was spent on things more product-development related. Only 75% of Complice tasks, but the programming tasks are more likely to have been eg 8h working on a single feature.
Founder of HR Partner here. I wrote the entire site from start to finish in about 3 months (from Nov/Dec 2015 to Feb/Mar 2016). Mainly working in the evenings (from 6pm to around 1-2am), after my usual consulting work during the day.
The site is around 22,000 lines of Ruby code, and has around 80 database tables. Not sure whether that is a normal rate of productivity for most devs, but as my interview explains, I am pushing 50 years old now, so my speed of development and my endurance is nowhere near what it was decades ago.
I did all the front end stuff as well for the public site, and the documentation site and the blog, status page site etc. About the only help I got was my wife to do the voiceover on the explainer video on the web site. (I had to learn After Effects to do the explainer video too).
I'm not featured on the site but I am a dev starting a company (http://www.askinline.com). I worked solo for ~2 months then asked a friend who I had been keeping updated on progress to join me as a co-founder. All up we've been going for just shy of 6 months and I believe we're just reaching the point where the product makes sense as a purchase decision for our target market. I say this based on continual discussions with early customers (in-network, pay nominal fee) and the fact that we're starting to land out-of-network customers.
You seem interested in the tech stacks. I have to say that when people have asked me what they should use I just tell them to go with whatever they are most productive in. Unless you're doing something to try out a tech or you're dependent on an outside system that demands a specific stack I wouldn't recommend going out of your way to learn too many new things while also starting a company. It's hard enough without adding tech concerns.
I (Apex Ping) went down the road of playing with newer more hyped tech like React etc, not really necessary for the project, and I can't say it saved any time (if anything wasted time) but still interesting.
AWS Lambda on the other hand has been a nice time saver aside from the fact that there weren't any good frameworks available really (with Go support) so I wrote https://github.com/apex/apex.
The only thing I would add is how many man hours the project took over what period of time.
I have to click on them and sort of guess if that site took 5hr/week for 10 weeks, or if it took 10 people working 10 hour days for 2 years.
Since you already ask for the tech stack, this would also help launch a dozen "which tech stack is more productive?" studies.
On a final note, I appreciate that you also ask about marketing. Maybe the marketing efforts and man hours can be summarized too?