Upgrading a sword using Rego craft magic?
We'll assume that the lesser sword has all of the basic bits needed to make a superior sword, so no further raw materials required. (Rego only makes use of what's available, it won't create anything additional, though it can skip steps that would normally use materials not present in the finished product.) You start with the basic ease factor for a Superior sword, 12. (City & Guild p. 67) Add +3 for the use of Rego craft magic. (Covenants p. 49, HoH: Societates p. 60) According to City & Guild p. 68, the average swordmaker can craft 2 standard longswords per season, but let's further assume that one attempting to make a superior one has a Craft skill higher than five, so that crafting a single sword takes one month or less. (Besides, we're starting with a functional sword-shaped object already made of all the correct materials, which almost has to speed up the process.) That adds another +3 to the ease factor. So the typical ease factor of the Finesse roll to make a superior sword from the base material of a lesser sword would be 18. (12 base + 3 Rego + 3 time)
However, if the crafter really doesn't know anything about swords - not normally a swordsmith, nor a trained combatant - he might take a penalty of an additional +3 to ease factor due to lack of familiarity with the item. OTOH, if he has a superior sword in front of him to study, or has Craft: Swords at a 5 or higher, he would lower the ease factor by 3 instead, reflecting his superior understanding of the task. (HoH: Societates p. 62)
Superior glass would be crafted the same way, referencing the same sections of the relevant books.
I personally rule in my game that tasks can be broken up into smaller steps, thus reducing the penalty completing extremely lengthy tasks in an instant. Glass making, for instance, was actually done in multiple steps in the medieval world, often at completely separate locations. One specialist location would produce the raw glass and ship it to customers. The secondary glassmaker would reheat the glass and shape it to suit. They don't have to start from sand and ash, they may not even know the full recipe or have a shop set up to make it, they only have to handle the final manufacture.
Likewise, a stone mason could throw up subsections of plain, slightly oversized construction with multiple uses of one spell; tie them together with another into a completely average, workable structure; and then use a third spell entirely to handle the fancy work and decoration. By breaking up the tasks, the ease factors become manageable. Not as sexy as throwing up a complete cathedral in an instant, but taking a couple of extra days on the project will likely yield a better built, better looking structure.
Of course, using Creo to conjure the glassware will get you the literal Platonic ideal of the stuff, but that's just cheating. Curse you, medieval philosophy!