Sometimes it's nice to discover that you are not the only one thinking about things in a certain way. From Kass McGann's "Women's Kirtle or Cotehardie or Medieval Dress" pattern description (my emphasis):
"At right is a reproduction of three characters from a 1350s copy of the Roman de la Rose -- two women flanking a man ... The women’s garments are similarly tight although they retain a high neckline. This is interesting to note because it is not the depth of the neckline that is villified in the accounts but the width.
Another illustration from the Roman de la Rose, reproduced at left, shows an interesting feature of this garment. This character is turned to the side, so we can see the silhouette of her body. Her garment is painted to look tight-fitting -- there are even some stress wrinkles on her back where it bends, showing us how tight the garment really is. And yet there is absolutely no indication of her bust. There is no roundness in her upper torso. There are simply no breasts depicted
There is a trend currently in the historical costuming community to create 14th century kirtles that lift the bust into an unnaturally high position, like that seen in marginalia of the Wenceslaus Bible, the famed “Bohemian Bathhouse Babes”. The argument made is that this lift is indicative of medieval bust support. However, the fact that the figures are not anatomically correct in any other way is ignored ... Why should we accept this bust position as fact when so much else in the paintings of this era is disregarded? We must be careful not to create a medievaloid answer to a modern fitting problem. This might answer the question of what should modern women do for bust support under their kirtles, but it cannot answer the question of what medieval women did without concrete evidence."