PSA: is our test... "inaccurate"?

The test has been released on October 9th and our inbox has already been flooded by hate comments.

TL;DR: If you are one of those who think that our test is inaccurate, just give suggestions as to how to improve it.
If you have no suggestions, please bear in mind that no test is perfect, and that your result can depend on a lot of factors. We spent almost two months working on it, consulting knowledgeable MBTI enthusiasts during development, and such feedback doesn't help anyone.

We see quite a lot of confusion regarding the interpretation of results, leading to a spread of unnecessary hate towards the test and the team. There are multiple factors to take into account:

• this is not a "personality test", but a cognitive functions test. It requires you to reflect upon your way of processing information (expressed through functions scores), not your usual behaviour.
• you have to take your time to answer the test properly: it requires lots of self-thought and introspection.
• you have to avoid "manipulating answers" or trying to assume which function a certain question is related. That is always a source of issues since lots of you may base your knowledge on stereotyped definitions of functions that don't match Myers-Briggs'/Jung's (on which these questions are based upon).
• tests are not 100% accurate. They are a starting point to base your research onto.
• anxiety or stress tend to bias your answers.
• taking the test on different moments is the norm to check how results compare to each other.
• the type you may believe you are may actually not be your real type (especially if you typed yourself based on I/E, N/S, T/F, J/P dichotomies).

Looks like most of the issues are related to:

• "reading behind the questions", unconsciously manipulating answers into something that completely shatters accuracy of the results.
• not thinking enough about your results. All the types are evaluated based on the strongest combinations of cognitive functions in different positions.
• taking the test to check if it matches with the type they are, rather than taking it with a completely reset mind (the answers don't match standard definitions, so you'll have to see how they apply to your cognitive processes).
• taking into account just the types rather than the functions themselves. The functions are what define your type, and the order in which they are classified is the most important factor to determine the most likely types. The result can get wrong especially if the scores are really close to each other.
• no interpretation of your results. It doesn't make sense to complain that you got two opposite types as your highest score (e.g. INTJ and ESFP) and blame the test not being accurate: they share the whole cognitive functions stack, just in a different order. That means the test actually caught that your type is within a restrict family of 4 types rather than 16 (you may be looping or gripping, or have overestimated the weight of your tertiary/inferior function).

This test is not meant to type you with 100% accuracy (studying the functions is the only reliable method) but as a tool to help you reflect upon your thought processes and eventually give you a direction as to how to type yourself.

If you think the test is inaccurate and boldly state it as such, please also give us concrete suggestions on how to improve it, otherwise it would be bad for the whole community.
Please keep the community healthy and safe for everyone.