The use of a household product as a release agent became popular simply because it works better, FOR THIS PARTICULAR APPLICATION, than the silicone based commercial release agents. It is easy to apply, can be used to temporarily block holes, is effective on all sorts of materials, and is easily cleaned up. It can be used to protect stock finishes from wayward epoxy; stock finishes which may be damaged by silicone sprays.
As far as bedding tight or providing clearance on the lug is concerned, the jury is, surprisingly enough, still out. When I did my first glass bedding job, (in 1966, at the age of 16), glass bedding was seen as a method to achieve the closest possible fit between stock and action. I was not yet involved in precision competition or rifle building and I bedded tight, with no clearance. I achieved decent results in spite of my ignorance. 10 years later, I had learned enough to know that I didn't know very much and I started paying attention to those who did. At the same time, I started taking an interest in serious rifle accuracy and benchrest shooting in particular. By this time it had become common practice to bed actions, especially Remingtons, with clearance at the front, sides, and bottom of the lug. A proper bedding job would allow the action to be easily removed from the stock once the screws were removed. That this formula was successful was borne out by the results on the line. That it wasn't perfect, was borne out by the occasional flyer. At the same time, high-power shooters were complaining that the round receiver of the Remington 700 and 40X was shifting in the stock due to torque. Some gunsmiths began to bed the sides of the recoil lug in an attempt to stop this shift in the bed. I cannot honestly say whether or not this was successful (or even necessary). Short range benchresters, always at the forefront of accuracy innovation, dealt with any tendency toward shifting in the bedding by gluing the receiver into the stock. This remains a viable and effective method of stock attachment and is undoubtedly where Accuracy International got the idea. The glue-in worked well enough that it was banned from use in Hunter class BR. Gunsmiths got around this by bedding conventionally (usually on pillars). They would then clean up the bedding, apply release agent to the receiver, and bed it again with epoxy only (no filler) while the bedding was curing the screws were torqued up and the action was not removed from the stock, though it could be if required. This fit into the letter of the rules if not the spirit. It was as close as one could come to a glue-in, without actually doing a glue-in.
Today, I still bed both ways, depending on the action type. Unless the screw goes into the lug, I always provide clearance under the lug. Unless there is no receiver forward of the lug, I never bed the rear of the barrel. On Remington 40X and Savage single shot actions, I always float the tang as well. The reason for taping continues to be ease of disassembly and re-assembly. I have seen winning rifles built using both methods. I would never be so arrogant as to claim my way of doing things was the very best way.