When I made the syrup, I based my work on a couple of assumptions, along with some work done by people like Mosher and other homebrewers over on the homebrewtalk board. The DAP is a just a source of nitrogen. Free nitrogen increases the rate of the Maillard reactions. The potassium bicarbonate is used to increase pH. PH affects the rate of reaction and the type of Maillard byproducts. More on that in a minute.
There is a lot of myth and misunderstanding in the homebrew community about how candi syrup is made. I am not an expert on candy making, nor do I have any firsthand info on how candy is made on an industrial scale, but I do have some insight on candy making on a stove-top.
Industrial producers "invert" their sucrose, probably because it's easier to handle as a syrup than as crystals. Once sucrose is inverted, it is no longer sucrose. It becomes dextrose and levulose (glucose and fructose). You can invert sucrose a number of ways, through any combination of heat, pressure, time, and acidity. Once the sugars are inverted, the candy maker would be able to precisely control the dextrose/levulose ratio. That is the end goal, I believe, and the fact they started with beet sugar is immaterial. They could have just as easily used cane sugar or corn starch, but I think because of the import tariffs they have in Belgium, it is more cost effective to use local beets than imported corn or cane.
Almost every homemade syrup recipe I found called for the addition of acid to invert table sugar. Sometimes they recommended quite a bit of acid. Acidifying the sugar does invert the sucrose, and that method works fine for making a light-colored invert sugar. You can also invert most of the sucrose in a short amount of time with just heat, like when making a simple syrup. So acid is not strictly necessary, since if you have enough heat and enough time almost all of the sucrose will invert. The problem with using acid to invert is that when trying to make darker syrups, the pH and dex/lev ratio becomes crucial.
Roasted grains are more acidic than base grains. Roasted grains are a product of Maillard reactions. So something happens in the Maillard pathway that lowers pH. The darker the product, the lower the pH. In my first trials with just DAP, dextrose, and heat, the pH on the finished product was below 4. How far, I'm not sure, since I have narrow-range pH strips, but the strips showed a strong and quick response on the lowest end of my strips. If I had added acid beforehand, the pH would have been even lower. The syrups also didn't taste very good.
Inspired by Dutch-process cocoa, I tried using an alkalizing agent added after making the syrup. The flavors became softer, more rounded, and overall "better." The pH of the corrected syrups was in the 5.2-5.6 range. A commenter on my original thread at homebrewtalk directed me towards some baking resources where they examined the effect of pH on bread. Alkalized bread browned differently than acidic bread. That gave me the idea to try adding the KHCO3 at the same time as the DAP. The Maillard reactions that happened in the 5.2-5.6 range tasted a lot better than the <4 reactions.
I also found that dextrose yielded much stronger/darker Maillard reactions than levulose. When using just dextrose, the syrups came out much too "flavorful" with a lot of negative roasted/burned character, and not enough of the softer "chocolate" type roasted character. When trying to make syrup with just sucrose, I got a lot of nice dark fruit/caramel flavors, but not enough of the roasted character. Through trial and error, I came upon a good mixture of dextrose to sucrose that yields a syrup with the proper amount of both types of flavors.