“but after studying many other ICO, I am convinced now that this is not just a short-lived fad, but a real new model to fund projects, and especially open-source projects related to blockchains.”
There’s been quite a few of us suggesting people jump into blockchain just to get funding for more important stuff (e.g. language-level tooling or libraries) they’ll produce as part of the coin offering they sell to investors. Plus, with the focus on correctness, it’s easier than ever to get the private sector to invest in high-assurance tooling. Problem is they keep trying to clean-slate everything when there's piles of work to build on with better benefits in the long term. And this work is hard even when doing simple things if you want end result to be widely usable. So, you better have a good justification for trying to redo decades worth of hard work on your own with some VC funding.
Anyone wanting a shortcut should build your language on top of capabilities of the SPARK Ada language  with this book . Your smart contracts get all the benefits it currently offers plus whatever extra you build plus whatever they build with the influx of money for Pro edition. Recent projects already add some pointer  and floating-point  safety to what they already mostly automate with that tooling. This is also a mature tool whose development and commercial use goes back decades .