It's far too late for this to have mattered at this point, but this is how I probably would have approached crafting. It eliminates the "you can't get that effect until you've done this a year" problem.
Start by assigning a percentage chance to generate an effect for EACH effect. The higher the ML adjustment for the effect, the lower the chance of success. The percentage chance get a bonus from the school.
As you successfully craft each effect, your percentage chance to succeed on that effect the next time increases, and earns you some school XP, which will go toward the school bonus.
This will cause many many failures when just getting started, and result in the plat sink that Turbine wants. However, if you REALLY want a Holy Burst weapon, you could try and try and try for the shard until you finally succeed at it. Crafting it once will make it more likely to succeed in the future. The high end ones will always be dicier than the easy ones.
Reward focus on one or two items and you may find people specializing in only one or two shard effects, which they can then sell or trade to others for their specialty shards.
Not-costly means of converting undesirable essences into ones that are needed. Sure.. make it lossy if you must, but not 2:1. 5:4 maybe.
Keep the progression in success rate at a very mild upward slant, so that if they ONLY do a single one, the bonus from the school doesn't make up for the ever-increasing costs, but that if they do several recipies from the school then the school bonus may even keep things nearish linear (for any given effect, but they need to do practice many effects to have it work.)
When merging the shards into the final product, allow optional ingrediants to increase success rate, lower effective ML, graft extra Prefixes/Suffixes (probably needing to be WITH the lower effective ML as well.), etc.
Ah well... not they way they're going, but I think it would have been more rewarding. Good to get it off my chest