RNG is usually based on current time, coming from the year all the way down to a microth of a second. It creates enough variance that it might as well be completely random.

Problem is people tend to focus more on the failures and ignore successes. For example, I can complain all I want about Hasty Touch repeatedly failing and never working, but if I look at the big picture over a long period of time, it usually comes out to around 70% success with Steady Hands.

Another issue is the size of the samples. Even 30 Garuda kills is extremely small and even then there's about a 2.9% of not seeing your weapon. You may say you've seen 9 summoner books and no scholar books and I can claim it's the opposite for me. This is also assuming that every drop has an equal chance of even dropping, which could very likely not be the case. Unless someone has the database from the game to confirm it, I honestly wouldn't be too surprised that the weapons that don't seem to drop that often from Garuda end up appearing more often from Titan.