As you now know, scale does
As you now know, scale does matter. What can be done at a small scale does not directly transfer to a larger scale.
The first problem you have created for yourself is using only one layer as the base. Glass has a surface tension which means that it tries to become 6-7 mm thick. One layer is only half that. As it thickens at the edges, it traps the air under the other parts of the glass, and as the glass softens the expannding air bubbles come up through the thin parts of the glass.
The design is the second problem. The weight of the border makes it even more difficult for the air to get out from under the glass.
The third problem is that you do not have a bubble squeeze in your schedule. You soak 10′ at 538C – not necessary. You do need a soak at or above 650C – this is the bubble squeeze. It is also the slump temperature, so you can determine what the bubble squeeze should be for your glass by what the slump temperature is. Your soak at 747C (tack fuse range) seems long to me, but you are the one who can observe and determine when the glass has tacked.
Fourth problem. As your separator is batt (kiln) wash, then you need to change to fibre paper for single layer pieces, as that allows more air out. Thinfire may be enough, but you can put it over 0.5 mm fibre paper for a smoother finish, or sprinkle powdered batt wash over the fibre paper and smooth it with a plasterer’s float or a piece of window glass.
Five. Look at the glass manufacturer’s website. Both Bullseye and Spectrum give basic firing schedules that work. I don’t understand why kin manufacturers don’t simply refer to the sites to give their customers good advice, instead of the pre-programmed stuff.