Joined: Oct. 2005
Say that a research university grants tenure to a 30-year-old scientist. The university is likely to keep that person for another 35 years, which could amount to on the order of 2.5-5 million dollars in salary, fringe benefits, and work space prior to retirement. Actually, since no one can make you retire any more, you could conceivably add another 15 or so years on to that, adding another 40% to those costs.
A popular teacher at a state university who pulls in 300 students into a large lecture course once a semester will earn their department somewhere around $200,000 to $300,000 per year for that course, depending on how much the university pays the department per enrollee at that course level. Let's say $240,000 at $400 per student. In contrast, an unpopular teacher who can only draw 30 students per intro session is only going to earn the department $24,000 per year for two offerings of that course. Over 35 years that's going to be a difference in $7,560,000 in departmental income. So a tenure decision is not a minor one, and a department will have a long time to regret a mistake.
Moreover, whomever a department tenures, the department is probably going to be stuck with for the next 35 years. A single person can quite easily turn a department from being a great place to work to a miserable place. Someone who carries their share of committee work is to be desired, whereas someone whose trials and tribulations creates committee work just drags everyone down. A good professor pulls in good students and adds to the glory of the department, whereas a bad professor scares the good students away. People have different strengths in different areas, so one hopes to get a good mix of strengths in a department. People say tenure is like a marriage (worse, because it's probably more like a plural marriage), but at least in marriages you can get a divorce if things turn ugly.
Thus a lot weighs on a tenure decision besides 'did this person meet the minimum requirements?'
Among the factors taken into consideration in tenure decisions (besides the numbers of papers published) are $$$ obtained in external funding; where were articles and books published; how many times have the various publications been cited; how many coauthors were involved in publications and grants; what were the candidate's contributions in multi-author works and multiple-investigator grants; how many masters, doctoral, and post-doc students has the candidate mentored; what senior people in the field but outside your university think of the person's research; professional awards and service; and teaching evaluations by students and fellow faculty. Along with other stuff. Shortcomings in almost any of these things could potentially doom a tenure decision.
To enlarge on the importance of where things are published: the humanities in particular have few ways of assessing the significance of a work (citation records for humanities works tend to be problematic and not comparable), so they have made a finely honed art out of assessing the reputation of the publisher of your book. This can arise, albeit to a lesser degree, in science as well. I would guess that a Regnery Press book might well be held against a candidate at a college- or university-level committee: it's sort of sending up a flag that you don't want to be considered a serious scholar.
(My apologies for some repetition of points that others have already made.)