If David's explanation relates to the clearance for the pin securing the sliding cap spring, then I am afraid David's explanation is at odds with the earliest examples that I have been able to find. So I am going to stick my neck out and say that I believe he is possibly mistaken. That is
unless, someone can identify an earlier example than the one described below, and can relate the 'feature', in some way, to the construction of the cap.
I refer to George Graham #5227 (p.184/185 The English Watch). This is one of the earliest cylinder movements made, after the cylinder escapement was developed by Graham in 1726. The watch is hallmarked 1727/28. Graham's earlier verge watches, and those of contemporary London makers, were characterised by cock feet that were much wider than the table and were heavily pierced. I have been unable to find a movement with a pierced wide cock that has the 'rectangular reserve'.
In describing this early Graham cylinder T C Cuss says ...
'The index plate and cock foot are engraved; it appears Graham ceased to pierce these out when he adopted his new escapement.'
As can be clearly seen in this photograph of the movement, the reserve is a cut out and is positioned, it would appear, to improve access to the pin through one of the pillars. It's position does not correspond to the securing pin at the midpoint of the cap spring.
A second Graham movement #5939 (sold by David) has the cut-out, but not corresponding to the position of one of the pillars. Again the cut-out does not correspond to the position of the securing pin for the cap spring. Similarly there is no correspondence with the securing pin for the cap spring in Graham #5983 - in this case there is a circular enlargement of the cut-out to accommodate the plate pillar, which is slightly offset from the cut-out. In Graham #6498 (1750) no correspondence with the spring pin and the cock has just a ghost feature which has been decorated to match the engraving of the cock foot. The second Ellicott example I posted in post #45 is from Ellicott cylinder #5052 believe to be ~1760 (#5141 is dated 1762 in The English Watch, p.218). It has the feature - now drilled for the screw securing the pillar. (Graham - do you think this is later?)
So my conclusion is that the feature was not introduced specifically to provide clearance for the securing pin of the cap spring. It may, where the feature provides an area of reduced cock foot height, have been used to advantage by the cap maker, but I see no evidence that it was introduced for that specific purpose.
In the earliest example I can find, as Chris pointed out ..
It actually appears to allow access to one of the pins
the cut out does appear to have a practical purpose. In later examples, it seems to me that the explanation may be no more than 'following that C18th fashion guru George Graham'

.
Willing to be proved wrong, if some one can provide a pre-1727 example that demonstrates a different explanation.
John