In my experience this is mostly not true. Most HR people don't know tech so what the HR screen devolves into is cramming your CV with acronyms so you make it pass the HR grep filter.
> You will want to interview a decent number of folks for every position.
If you're bringing in more than five people for onsite interviews then (IMHO) you're doing it wrong. You're wasting your time and theirs and it's clear to me that you don't know what you're looking for.
At least Fred advocates phone/video screens (of 15 minutes) although he suggests 6-12 onsite candidates.
> Many employees don't know how to interview and you should teach them the basics ...
This is true but I would go further: most people shouldn't interview. It's arguably a talent not a skill. At Google, it's viewed as every engineer's responsibility to do interviews and I disagree with this. Many of the stories of bad interview experiences can be attributed--at least in part--to people who just don't have the interest or talent in interviewing.
Anecdote: in one quarter I once took 10 people to lunch as part of their interview slate. In those lunches I just talked about the company, answered their questions, etc but didn't get into any technical discussions.
Of those 10 I predicted 8 wouldn't get offers, 1 would and 1 I was borderline. Actual results: 1 offer, 9 no offers putting me at 9-10 out of 10.
> If you connect to the candidate on LinkedIn, you can quickly figure out who you know that knows them.
What an incestuous world the NY/SF tech scene must be if this is true. Well it might be true if you're a veteran VC. I can't see it being true for anyone else.
Otherwise this is all straightforward good advice although it skips the most important step: how to determine if someone is a good fit and competent but that's a topic we'll probably argue about in perpetuity.