How were the prickly pear pouches prepared and how may they have been used?

This question is partially answered in an encyclopedic study of Lower Pecos prickly pear pouches done by Eric J. Brunneman, then at the Witte Museum, for his master's thesis at the University of Texas at Austin. Brunneman approached the problem of prickly pear pouches by summarizing all of the ways in which prickly pear was used and/or consumed by various Native Americans, ethnographic references to such artifacts, descriptions of archeological specimens, and replicative archeological experiments. Of particular interest are his citations of Spanish accounts of prickly pear pouches in Texas and northern Mexico that are relevant to the Lower Pecos as well. For example, in the late 17th century, Alonso de Leon observed women carrying net bags filled with prickly pouch canteens. He also noted that ground mesquite beans were stored in split prickly pear pads. Brunneman's main concern was, however, replicating the archeological specimens resident in Texas artifact repositories. The analysis concentrated on 52 prickly pear artifacts, 29 of which were pouches, from the Lower Pecos. His replicative experiments showed that the best raw material is what might be called the middle-aged pad - not the oldest interior pads that are hard and brittle or the outer juvenile pads that tear too easily. He scraped the spines from the pads in the field without tearing the epidermis (using flint flakes and a scraper that he made), bundled them, and tied the bundle with sotol leaves for transport, following the description of a similar packet found at Baker Cave. In the lab, he cut a slit in the perimeter of the pad, beginning near the stem, and extending about a third of its length. He then split the pad open by inserting his fist into the incision. He did not scrape or cut the interior of the pad. Eight of the pouches were stuffed with various foods - onions, fish, lecheguilla hearts, prickly pear tunas - and tied shut with sotol thongs. Apparently, the tying was essential to keep well-stuffed pouches from buckling and spilling their contents. A cooking pit was dug, lined with limestone, and a fire built to heat the rocks. The stuffed pads were then cooked on the hot rocks, turning every 15 minutes to avoid burning the pouch. Awful as it sounds, "the mucilaginous interior pulp of the prickly pear pouches acted to steam the contents..."The external burn patterns matched those on many of the archeological specimens, thus indicating that this form of broiling was one primary use of the pouches found in dry rock shelter deposits. Although Brunneman's focus was cooking in prickly pear pouches, he made some findings that may be relevant to the production of pouches for other reasons as well. First, as mentioned above, the best pads were mature but not too old or too young. Although the thorns were scraped from the pads, the epidermis was left intact and the interior was not pulped. Baking the pads seemed to retard rot. Brunneman has a great deal more to say about prickly pear pouches than I could summarize here. His thesis "An Artifact Analysis of the Prickly Pouch: A Museum Studies Archaeological Analysis in the Lower Pecos River Region of Southwest Texas" (1988) is available in the Perry-Castaneda Library and at the Texas Archeological Research Laboratory, both at the University of Texas at Austin. I suspect that the Witte Museum also has a copy on file. I would be very interested in hearing from anyone who has experience in replicating pouches for other uses, such as canteens.