It got patented but had long expired by the time that crew designed (not discovered) their transistor.
If you've ever tried positioning a cats-whisker on a piece of galina to make a diode for a crystal set, imagine positioning TWO of them correctly so the interaction makes a working point-contact transistor - and doing this half a dozen times -= with them ALL working - to get enough working stages of amplification to have an effective radio.
I think what happened is that the gain was too low and the construction too chancy, and people believed that it was bogus.
Remember that - except for perpetual motion machines (where they demand a working model as a way to keep the crackpots out of their hair) - the patent office is only concerned with novelty, not with whether the thing actually works. (That's between the inventor and his licencees.)