So you are of the opinion that using strong pull-ups, there is no need for software debouncing?

Bouncing varies from switch to switch, and also varies upon each press and release. Just look at the oscillogram you had posted. The time is in milli seconds and the microcontroller operates in micro seconds. In the upper waveform, micor might read 3 switch presses if the there was no software debouncing. Also upon release, it will read another switch press. (Of course, it depends upon Vih and Vil levels of the particular micro used).

However, if you don't want to use software debouncing then you can use RC filtering. The best way to check is to wrtie a routine which detects and counts the number of switch presses and displays on LCD, without implementing software debounce. You will realize how necessary it is to debounce!