While I agree the 2# per inch projection can get you into trouble I disagree on the remedy. I find if I tiller a bow to max draw weight all the way up the draw lenght, and especially if I "work" the adjustments in as conventional wisdom suggests by pulling an arbitrary number of times at each station, each inch of draw say as I tiller it out, I finish with a bow drawing the poundage I wanted and making cast like a bow 5# or more lighter. IMHO you are better served by drawing the bow as little as possible in general, and only enough to find flaw with the tiller. Then proceed to projected draw weight at a particular lenght, eg. not the full draw weight as you continue to adjust tiller. But you do have to allow an extra 5#, on average say for a 50# to 60# bow at 28", of lost draw weight from set. So if I want a 50# finished bow, I project back 2#/" from 55# for my max weight at each inch of draw. The set will usually come when you get to about 22" to 24" of draw, again assuming a typical design, moisture control, all the normal stuff. I try to get my tiller dead nuts before I get to this point, to try to minimize the lost draw weight but more importantly the lost cast. Maybe not the best advice for a fellar just trying to get his first good bow, or having just made a few. But after you've made a dozen good shooters and you want more, that last little bit of performance, it's something that will open your eyes when you get it right.
To put it another way, if it were possible to cut a bow out to final dimensions, perfect tiller having never drawn the bow, then when you first work the bow it may drop a few pounds of draw weight to set but it will make cast like a bow 5# or more heavier.
When I first started out, after a couple three years and as many dozen bows, I was amazed at some of the chrono numbers I saw posted for bows very similar to mine. Eventually I figured this out. I think most bowyers follow the same path, employ the same technique without realizing, simply by developing a better intuition for tillering from experience. This having the net effect of not needing to draw the bow as many times, or to as high a draw weight/strain level to get it "right."