If you want to implement something quickly and even more so if you want to maintain that speed in the long run, then there is one thing that will turn all the t

If you want to implement something quickly and even more so if you want to maintain that speed in the long run, then there is one thing that will turn all the things the article talks about into mere noise - design things properly. Especially data models. After that comes implementing things probably, i.e. no shortcuts and hacks. The things the article talks about are not bad advice but I do not think any of them are critical for being fast besides getting and incoperating feedback as early as possible. This is of course not limited to feedback on an implementation, you of course also want early feedback on your design before you even start implementing something.