I don't suppose you have any statistics from other takers of the test how the bars relate to the development of those catagories?
"The test" is a loose term. If you want to look at experimental data from the original validity tests you're talking about the MBTI Form M. Internet tests are not Form M. Most copy the questions (which is plagiarism, but the Myers-Briggs Institute hasn't traditionally sought any damages on that count) though not the individual question weights, and many others are new but completely untested. Since it is the only type theory test I'm aware of with any compelling experimental evidence supporting it, lets look at the official MBTI.
If you don't want to dig through research papers I would suggest skimming
this. Many of the statistically significant factors are referenced in the 3rd edition MBTI Manual, which is an excellent place to start for a thorough look at the theory and its history.
One concept that comes up repeatedly reading through the manual was that one of the most difficult parts of writing the indicator was that this is a type model rather than a trait model. That means that for validity you must write it such that the results can not be modeled by a Gaussian. If the results fit a Gaussian it means the bulk of test takers would end up at the middle and assigning them a type would be meaningless. The best possible result would be a Gaussian halfway through each side of the scale. To achieve that end very similar questions are asked dealing with the same subject matter. The test taker's preference ideally comes through by answering that question significantly more times in one direction than the other.
The weakness of the instrument (though administrators of the MBTI rarely think of it as such) is that beyond classifying the individual into a generalized group, the test does nothing. Everything else is dependent on the untested factors, such as environment of upbringing, life choices and educational exposure.
All that to say, the indicator was never written to determine the degree of maturity or function development in an individual. I have doubts that such a test could be accomplished with multiple choice, much less a forced choice format. Hopefully it will put you on a path to better self knowledge, but it claims no other purpose.