This is not so much a question as a cautionary tale that may be encountered if you follow the datasheet recommendations without applying your particular details...
Problem: reset_b is kept low during debugging, causing "cannot halt processor" and a failure to connect to the target.
My reset circuit follows the one in figure 37 of the datasheet (attached) for the MKE18F.
The problem, apparently unrelated to the MKE18F, is that the external reset module includes a pushbutton debounce 'feature', where it senses if reset has gone low for more than 1 uS, and if so, holds it low for an additional 150ms to debounce an external pushbutton switch.
The problem is that this 'feature', combined with the RC circuit shown, means that when the reset module releases reset, the RC keeps it low for more than 1 uS, causing the reset module to trigger it's 'debounce' feature and pull reset low again. Reset is never released. Omit the cap, and it seems to work.
Unanswered question: somehow, it worked sometimes.
Hope this helps.