Like any commercial company, Mathworks has existed for a finite time, and has finite resources to program, test, debug, and document software. As is common for companies, it prioritizes writing code that will benefit many people, or that will open up a new category of software use (extending it to more topics). It employs people who know particular fields (for example, electrical simulation) well, and so know what kinds of tools are in use in those fields. It accepts feedback from customers as to what more they need: the more customers that write to them asking for a particular addition, the more likely they are to implement it.
Thus, if there is a particular field that often uses pair-wise heat-maps, then Mathworks would appreciate if someone would send them a use case showing what should be added and what benefits it would have for people in a field.
Mathworks also hosts their File Exchange, which permits user contributions. I suspect you are referring to that when you talk about putting in packages with no literature or any support, and not standardized or structured.
Standards requiring code to be structured a particular way, or present a particular "user experience", or must meet particular documentation standards, or must meet particular unit test standards -- all of those act as barriers to volunteer contributions.
Mathworks is, for example, happy to have a student contribute a bit of code along the lines of, "Hey, I had this problem, and this is how I solved it". Such contributions get the people involved in the community -- and if maybe the earlier such contributions from a particular user are not very polished, the later contributions will probably be more polished. A lot of the contributors to the File Exchange do not write English well, so if the contributors were required to write extensive documentation in English... a lot of them would not bother.
About the only alternative would be if volunteers were able to contribute whatever code they wanted, but that Mathworks then took the code and polished it up and standardized it before releasing it to other users. That is unrealistic, though: there are more than 40 thousand community contributions in the File Exchange, and it would require a lot of time and investment to rewrite all of the volunteer contributed code to meet some standards.
The File Exchange is free software, users making their works available to other users. It is not some kind of professional marketplace where people are earning money from the code they contribute and so could conceivably be required to meet minimum quality standards.
Perhaps you are thinking, "Oh, but look how many quality free contributions there are in Apple's App Store! If Apple can enforce quality standards, surely Mathworks can as well !! " . If so, then such a thought would ignore that Apple charges $99 per year to authors of contributions to the App Store, and would ignore that Apple takes 30% of the app revenue (including any in-App purchases.) Apple makes a heck of a lot of money from their App store (over $1 trillion per year; https://techcrunch.com/2023/05/31/apple-touts-1-1-trillion-in-app-store-commerce-in-2022-with-104b-in-digital-sales/ ); they can afford to put resources into it.
