-not much time has passed since the first use of language (by prehistoric people) to this day, so it can be assumed that only a negligible part of the possible mental calculations/connections has occured
Regardless of when language emerged (plausibly 50,000-200,000 years ago), we can probably agree that only a negligible part of “possible mental connections have occurred.” However, this in itself does not seem a compelling reason to worry about a hypothetical mental illness that we have never seen before.
-there is no direct survival bonus through ability to think in complicated manner; on the other hand there is arguably an cost-effective logic in disabling great freedom in self-examination
People don’t live merely to survive: we’re hardwired to propagate our genes. If you cannot think abstractly and articulate your ideas well, you will have difficulty attracting a mate. People who have disabled their ability to examine themselves will be quickly eliminated from the gene pool. Hence, it seems unlikely that such an illness will occur because it goes against how natural selection has shaped us.
At any rate, it is just my guess—there are so many unknowns about the mind that this may too be impossible to actually happen. One reason why it would be unlikely is that, ultimately, if so grave a danger was built-in a system, it would make more sense to never allow as an option the expansion of ability to think in the first place.
This reasoning seems to rely on the assumption that the mind was designed by some kind of agent. Who do you think is deciding whether it “makes sense” to allow an expansion of the ability to think? Our best theory is that cognitive expansion resulted as a series of mutations that improved the ability of our ancestors to survive. One does not need to appeal to the fact that “Day Zero illness” does not “make sense” to argue for its implausibility. It is implausible simply by the fact that it is a priori highly unlikely for any novel previously unobserved phenomenon to exist in the absence of a very strong theory that predicts it.
I made tulpas because I was curious about the phenomenon. I did not find the creation process difficult. I thought for a long time about how to make tulpas useful but the best application I could find for them is possibly as a way of training an internal random number generator. I imagine they would be useful for fiction writing as well.