This is an excellent example of fud created by security pundits wanting to say "actually, that's not secure". If you read that link carefully you will find that the problems discussed are not unique to digital redaction (size/length of the redaction, height of the redaction, etc). The accepted answer also clearly stated that if you use pixel-replacement (this is what DoD calls it) you're fine. Further, there have been plenty of times I've seen a redaction-by-marker break down when the underlying ink bleeds through in photocopy.
NSA published guidelines in 2005 for how to safely redact PDF (as well as word) documents. The corresponding physical redaction guidance begins by suggesting that the inspector physically remove the information to be redacted with an xacto knife. They redact and release thousands of documents, if nobody has reversed their redaction process yet then it's good enough.
Also note the answer that points out underhanded ways to leak data.
You then go ahead to recommend just the thing that gp was making fun of: printing and physically altering the document instead of using a software tool.
You should re-read my post. I said that there were standards for digital redaction which are considered sufficient and equivalent to physical redaction. Break. If you still want to manually redact... iow, manual redaction is a waste of time, but if you have time to waste here's an effective way to waste it.
May is a pretty clear clue that the security "advice" which follows is fud. Your house may flood even if it's on top of a high hill, buy my insurance product.
See this for some food for thought on some of the ways digital alteration can fail you: http://security.stackexchange.com/questions/126932/how-secur...