One can find scattered throughout the mountain literature references to the height of a mountain above its "base." Such references are often used to describe and compare the visual impressiveness of various mountains. The trouble is that there is no real well-defined notion of the base of the mountain. So these numbers, and the resulting comparisons, are dubious. Also, the "height above base" number does not capture visual impressiveness, since it does not take into account steepness.
Prominence has some correlation to the ill-defined notion of height above base, but prominence compares peak heights to saddle heights, not bases, and can (especially for high-prominence peaks) use a saddle that is a large distance away from the peak. Nonetheless, prominence aficionados are usually interested in the visual impressiveness of a peak, and that is often part of the motivation for investigating prominence. So it was on the prominence@yahoogroups.com mailing list that discussions arose about such notions which led to the creation of spire measure.
Bob Bolton originated the idea of having a measure which correlated better than prominence does with visual impressiveness. He came up with the seed of the original idea, at a time when I (David Metzler) had despaired of having a better "impressiveness" measure than prominence. Bolton suggested measuring (height * slope) from the "base" of a peak, in various directions, where the base in a particular direction would be defined to be where (height * slope) was maximized. Spire measure grew out of this suggestion, but is a different concept in important ways.
The first problem with Bolton's suggestion is that it is not bounded; it becomes infinite for a vertical cliff, no matter how tiny. Edward Earl proposed a bounded version of Bolton's (height * slope) formula, namely (height * slope/(slope + 1)), which still is at the heart of spire measure. The other major difference is that spire measure takes into account (integrates over) all of the surrounding terrain, in a way which naturally weights closer points more strongly. This avoids the ill-defined notion of the "base" of the mountain. Edward Earl and David Metzler took these ideas---a bounded version of height*slope, integrated over all surrounding points---and developed spire measure. (Here is a detailed description.)
Spire measure, however, is not sufficient for making a "Top N" list of peaks, since it does not automatically exclude subpeaks. Hence Earl and Metzler developed reduced spire measure, which takes into account independence as well. Most of the lists on this site are based on reduced spire measure. Earl and Metzler also developed the use of spire measure-based ruggedness, which also appears on this site.