This is a follow up question from here.
Since I'm allowed to delay the distribution of the binary and source code as long as I provide the source code eventually. This makes me wonder if that does not defeat the purpose of the GPLV2 completely. If that is true, I could choose an arbitrary time when I release the source. So, I might change GPLv2 code and release the source eventually, years after the binary when my program does not have any business value.
I'm asking this question because I'm annoyed that Apple releases their open source component (some which are licensed under GPL) deliberately after their binary release.