Here are a couple of interesting scans from books I have on the subject. The first is from William Gamble, "Music Engraving and Printing," 1923, pg 131:
He avoids going into an exact method stating "It is more easy for the beginner in music engraving to have a definite amount to give for the spacing, but the experienced engraver does not need such aid." To me this seems like he is going by eye and not really thinking in terms of a mathematical ratio. When he does give mathematical spacing advice, it is just along the lines of "if the spacing is set for a crotchet, one more space must be allowed when a minim is encountered." Quite frankly, I don't find this terribly helpful since we don't know the amount of space he's allotted for the crochet. Is he really saying regardless of how many spaces a quarter is allotted, a half is always one space more? I find that pretty hard to believe, but instead is conveying the point that obviously a half needs to be allotted more space than a quarter, but probably not twice as much, or he would have likely said so.
Here's Ted Ross on the subject too, from his 1970 book pgs 76-77:
In contrast to Gamble, he actually dictates an exact formula. If we examine the ratios, eighth to quarter is 1.4, quarter to half is 1.36, and half to whole is 1.53. I'm assuming these must be close to what Sibelius is using in the "Ted Ross" spacing algorithm. Looking at the "Sibelius - Ted Ross" example I posted above, to me it seems like the half doesn't get enough space compared to the quarters and eighths. If they are using these spacing values, that seems to make sense as the quarter to half ratio is only 1.36. There is a lot of personal preference here, but remember 1:1 means all notes get the same spacing, so anything sub-1.4 seems like the note values aren't being represented as clearly as they could to my eye.