When people read "V compiles itself in less than a second", they expect V to compile itself in less than a second for the general case (not just one very specific case). When people read "as fast as C", they expect V programs to have the same runtime as equivalent programs directly written in C. When people read "no null, no globals, no undefined behavior, immutability by default", they expect there to be no issues with this 99% of the time. When people read "Despite being at an early development stage, the V language is relatively stable", they don't expect this number of issues. I could go on, but the V documentation is filled with this.
In my opinion, if you keep going in this direction, no one will ever take V seriously. You might get some GitHub stars, you might get some inexperienced language developers helping you out, but if people can't trust V's claims, if V's claims continue to not hold up, no one will use V anything more than casually.
The simple solution to this is to walk back your claims and preface everything as unstable which is more accurate. Your GitHub description should make it obvious this is an experimental language with many issues. You should only list features in your README which are true and work for the average or general case. Any other features can be listed somewhere else as in progress or testing. Toning down your marketing is all it takes to earn back some trust.