Programmer Number Influence Memory Consuption

I've noticed a strong correlation between the number of programmers developing a program and that program's memory consumption. If company has 10 programmers, then its product would take 50 megabytes RAM, if it has 100 employees that would be 500 Mb RAM. For example, Mozilla has more than 1000 employees, and Firefox requires several megabytes RAM. Dillo has just a few developers and its memory consumption is below 50 Mb. Same appears to be true for operating systems: early Windows NT has a small development team, which appeared to have very informal development practices (judging from a lot of swear words in comments), so Windows NT required just few megabytes to run, while modern Windows 10 requires gigabytes.

>>16,17
I'd say that things like SQLi, filesystem access, shell injection etc. are a different beast from buffer overflows. their're are less about low-level vs high-level languages but about low-level vs high level interfaces between different systems or languages. this isn't something solved by a safe language (no amount of compile-time type checking, bounds checking or anus checking will detect that - unless you bake things like SQL access into the language itself, but that carries a whole set of different problems), it's something solved by a well thought-out abstraction