The project got first customers in 2005 so it is afloat 13 years so far. Until 2014 it was a side project. Since then I am working on it full time as a solo founder and developer.
Initial version was pretty simple so I was able to do it after official work hours and at weekends. Thus I didn't need any funding at that time.
The project has definitely over grown single person model now so I am actively looking for partners, cooperation, investors, teams, etc.
So are couple of advices:
1. If you can do it initially as a side project - consider this option first. Could be hard but will minimize negative budget impact and other risks.
2. The project shall reach the stage of minimal viable products in not more than 9 months. For many reasons.
3. You should do it in modular architecture: if first one will not take off you will have ready to use components for the next one.
This is actually really great advice! I sold my previous side project and still had lots of reusable code to use for https://tinytracker.co
The question was on different angle: can single person compete with the whole IE team at MS to achieve anything viable. And that required the engine to be present to try by others.
Initial version was very basic HTML renderer named HTMLayout that time. Made site for it and published couple of articles about it. These allowed me to gather requirements and estimate interest/market for it.
After that modules of HTMLayout were assembled into Sciter, with additions of CSS and scripting.
As of 9 months...
I've participated in many projects, successful ones took 9 months from first PRD to alpha/beta stage. Longer projects, as a rule, were less successful. Team lost the steam, etc.
YMMV of course, but 9 months to create something, seems like is embedded into human nature. I cannot provide any formal proof, just a feeling supported by 30 years of experience in the business.