Although i understand some frustration with parts of the game engine and i equally understand how and why people would want the code, i'll explain why (according to me) it's a poor idea (and this is regardless of wheither Marin understands the engine or not, but i reckon he understands enough about it).
A) if parts of the code would be given, then a programmer familiar with that language, can understand certain aspects of the game better than others. This creates an unfair advantage to some users (mind you, i've got the knowledge of several programming languages, so i know what i'm talking about). If something would be released, it would have to be pseudo code, simplified to a lvl where everyone can understand what is written (which probably will be a lot longer than it needs to be).
B) if you hand out (parts of)(pseudo) code, then the users will be able to train the best players to fit the system. right now, nobody knows which is the best tactic. The (isolation) tactics are still misunderstood by vast majorities of the community. by trial and error (or by analysing data from other teams, which involves teamwork because you need to work with other managers to correctly understand the data you are receiving ==> community working together and exchanging info) one can understand more and more of the game engine. However, if you know in advance how the game will turn out, then where is the fun in playing the game? It's not fun for the team who would be losing all the time. And if you continuiously stomp your opponent, it gets rather boring too. Cause giving up (parts of)(pseudo) code.
C.1) to achieve a goal, you can change many things. If (parts of)(pseudo) code were handed out, then Marin would constantly have to change parts of the game engine to guarantee a continiously changing game. this brings in the risk of having ending up with an extremely unbalanced game. Making minor changes to a GE can have huge impacts. it trickles down to tactics, but salary aswell. all of a sudden, your players might be more or less salary efficient. and you wouldn't be able to plan it in advance.
C, 2) to continiously guarantee minor changes to the game, one can change the GE code. or one can change the salary slightely, making certain players more cost efficient that others (or to balance certain players out). I prefer the salary changes (which one can somewhat forsee to a certain degree) than GE changes (which nobody could forsee, meaning that your team might be champion material one season and could be facing relegation the next).
Consequences:
Huge new interest and debate among user with knowledge in programming, or mathematics, or with big game experience. New thriving research programms are started thanks to the input of the new knowledge.
Interest from new users in this innovative way of managing an online game.
By giving input in the suggestions forum, you can also achieve these goals. if the suggestions are considered good, you could ask for someone from the community to provide a pseudo code. By doing so, all the BB's would have to do was to incorporate the code into BB (provided it's possible). there are a lot of ways to help them out, but it all starts with a good suggestion.
So i understand your request. but the negatives far outweight the positives of your suggestion. i prefer slightely unpredictable nature of the game (like in the NBA, where the less quality teams can beat the top teams on any given day. Is it likely to happen? Not that often. but when it happens, it spices things up! keeps things interested! and for that reason, i don't support the suggestion, but that is just my personal oppinion)