It does not matter what system, methodology, programming language or even RTOS one uses, whether it works or not is down to the programmer

excellent, may I add that ever so many faulty designs is due to someone believein that using a given "system, methodology, programming language or even RTOS" will cure all woes, when it, indeed, is "down to the programmer".

My experience is that most such happen when a PC programmer is tasked with coding embedded and try to force the PC methodology down on the poor little '51. I have seen projects that did not work because of things like "I use the huge model because I can not be bothered with memory selection" and "using printf in an ISR is not a problem" and "using malloc saves memory". To save myself from Donalds ire, i shall not mention the other thing some of them do. I shall stress that there are, indeed, PC programmers that embrace the effort to make the transition and do very well. However, when a company with no embedded background hire "a" programmer to do a '51 project, the likelyhood of the above happening is large. In case anyone wonder, I have done embedded my entire career, my first project embedded a PDP-8

Erik