Chau,
I also am not an expert on any of these matters. However, I would like to offer up the following comments and observations:
1. I, too, experienced many of the principles you observed when I attempted to make long (20-24 hours), room-temperature fermented doughs. In my case, I made straight (non-autolysed) doughs. I reported on my results in the opening post in the thread at
http://www.pizzamaking.com/forum/index.php/topic,7225.0.html. As noted in that post, I concluded that for such a dough to succeed in my high room temperature environment, it was necessary to reduce the hydration value that I would normally use for the dough formulation in question by about 5% from the rated absorption value for the flour I used. I did this in the belief that the protease enzymes were destroying the gluten structure and causing the water in the dough to be released from its bond. This latter effect also made it necessary for me to re-ball the dough a few hours before use to improve the dough strength. Ideally, I would have preferred not to have to do any re-balling, as I so noted in the above-referenced post.
2. The protease enzymes are temperature sensitive. They will be denatured and rendered ineffectual (non-functional) when the temperature in the dough reaches 60 degrees C, or 140 degrees F (see
http://www.classofoods.com/page1_7.html). So, for all practical purposes, the protease enzymes (along with other enzymes) will be destroyed during baking of the dough. The protease enzymes are also pH sensitive and can be rendered non-functional at certain pH values. However, these values are unlikely to obtain for most normal doughs, and certainly those that are used with commercial yeast rather than natural leavening systems. Finally, the protease enzymes are salt sensitive. That is one of the reasons why salt is omitted from a classic autolyse--to optimize the softening effects of the protease enzymes and improve hydration of the dough (and shorten dough production times). On the other side of the issue, it might also help explain why higher than normal salt levels are used with doughs, like Neapolitan-style doughs, that are to be subjected to long room temperature fermentations. The salt, in addition to helping strengthen the dough, will keep the protease enzymes in check so that they do not materially harm the gluten structure and release the water from its bond. As I see it, so long as the protease enzymes are not functionally disabled, they will continue to work to take apart the gluten structure. Given enough time, they can literally destroy the dough. I tested this thesis successfully for both commercially leavened doughs and doughs with natural leavening agents.
3. About a couple of years ago, when my interest in autolyse rest times was piqued, I looked at the dough recipes that Professor Calvel, the father of autolyse, put in his book
The Taste of Bread to see what duration of autolyse rest periods were used. I reported on my results in Reply 15 at
http://www.pizzamaking.com/forum/index.php/topic,3220.msg74624/topicseen.html#msg74624. As noted there, the longest autolyse rest period was 30 minutes, and that was for amounts of dough of over 70 pounds. Some members, including djones48, experimented with much longer autolyse rest periods (see the thread at
http://www.pizzamaking.com/forum/index.php/topic,7770.msg66722.html#msg66722) although in his case he used an overnight cold autolyse and he was looking for more sugar production than improved hydration. I believe that John (dellavecchia) also conducted some experiments with autolyse rest periods far in excess of the autolyse rest periods contemplated by Professor Calvel (the Suas methods that John has cited).
To be honest, in the final analysis, what is more important in my opinion than trying to fully grasp all of the phenomena and principles at play is whether the final results are acceptable and whether they can be reproduced on a consistent basis. It is much like how a child, like my 9-year old granddaughter, can use her iPad with remarkable facility without having the faintest idea as to the technology behind the device. By analogy, it is like presenting a dough formulation with precise values (like baker's percents/weights/volumes) and a detailed set of instructions that just about anyone should be able to follow. It would be a different story if one were tasked with creating a new dough formulation for a particular application. Then, it certainly helps, and may even be necessary, to understand most of the phenomena and principles that come into play.
Peter
