Having to first load a template to the pad, and then open the iptc field and apply it, is obviously slower than what I'm doing, and one is no more or less error-prone than the other. Indeed, I would think that the pad method is more error-prone, because you don't know for sure what template is loaded to the pad without checking first.
The apply stationery pad method is fine if you don't need to switch among several different templates on a frequent basis. My workflow requires switching templates on a frequent basis. Having to constantly open and close the pad every time I want to change templates would slow me down significantly.
My method accomplishes the same thing without ever having to open, load, and close the pad, then reopen the iptc field and apply the pad. I never have to close the iptc field as I go from photo to photo. That's not a pseudo-thing, it's a time-saving great thing! It will be faster for me to work around the loss of the append by copy/pasting the caption to a notepad, applying the template with the load button, and then pasting the caption back in, above the info that the template put in. That will be faster than changing the pad every time I need a different template. sigh.