Up front, I'm not an Epic level person, so I'm not the target audience. Please take my advice with a heap of salt.
However, overall this process looks overly complex - lots of math and options that make it ripe for abuse. Personally, I would think it would be better to simplify it and treat it sort of like the Warlock's Mystic Incarnum - just covering a set of spells over 9th level. Rather than base the number of Epic spells that can be cast, limit it to 1 per long rest (or maybe, 1 per week), but have the # of seeds known either at a set rate or controlled by spellcasting modifier. Also, set the DC to 8 + PB (or x2 PB, as this would be "Expertise Casting"?) + spellcasting modifier, like other spells, rather than calculating the DC by how the spell is built. I'd strongly consider giving "Epic Spell Points" in a small number to allow for the spell seeds to be customized (maybe start at 1 and gaining 1 per level past 18th or 1/4 class level and work upwards or downwards from there?).
I'd keep the idea of seeds, but simplify them to a degree with a "at higher level"/per Epic Spell Point effective section for each. Personally, I'd have all the seeds have a base expensive material component (something like 2,000 gp x spell level or Epic Spell Point of a specific item). Move the "factors" directly to the seeds, set out under the "at higher level"/per Epic Spell Point section. Since it would be more fixed, I'd consider tweaking Casting Times, Range, Targets and Duration for each of the seeds.
It might be worth creating a master table for each of the above, with each seed having a default setting, and allow a sliding dial - tweak one component down to move another up (possibly with an additional material component cost?). Perhaps, something like this... (blue is the "default", no seed should start in the red)
Casting Time | |
Glacial | 7 days |
Daily | 1 day |
Long | 1 hour |
Slow | 10 minutes |
Moderate | 1 minute |
Standard | 1 action |
Quick | 1 bonus action |
Range | (reversed ranges so "Self" is a regression and "Extreme" is an improvement |
Self | Self |
Touch | Touch |
Radius | 15 foot radius |
Short | 30 feet |
Medium | 120 feet |
Long | 360 feet |
Extreme | 1 mile |
Target | |
All | All targets in range |
All, medium burst | All targets in a 15 foot radius of the target square |
All, small burst | All targets in a 5 foot radius of the target square |
Three | Three targets, within 15 feet of each other |
Two | Two targets, within 15 feet of each other |
One | One |
Duration | (for epic spells, ALWAYS requires concentration except if Instantaneous) |
Permanent | permanent/100 years |
Year | 1 year |
Extreme | 1 day |
Long | 1 hour |
Medium | 10 minutes |
Short | 1 minute |
Instant | Instantaneous |
Combining seeds should effectively raise the spell's level - I'd suggest +2 levels/Epic Spell Points per combined seed. When combining seeds, the most restrictive Casting Time, Range, Duration and Target should be used.
I'd conclude with a section that gives the DM control on allowing the PC to come up with ways to possibly mitigate or enhance the spells for more "epic" effects or to enable casting certain spells more reasonable. Things like questing for supernatural components or places of power to negate/reduce costs or shift Casting Time, Range and the like. Possibly allowing for special circumstances to modify seeds beyond the ranges listed in the tables.
No formula, so that the DM can keep things from getting out of control and ruining the game, but still make using the ability something enjoyable for the player and the group.