Hmm, so the benefit from spells and phemes stack? Is there a limit? I never really understood how the spells and phemes operated during my playthrough (stacking, etc.) but from reading the forums here, the following seems to be how I gather it works:
Quick math -
6th finger adds 2 to Enchant (and Forge, but not directly relevant), and has difficulty 3. Since Int + Enchant is the roll and Int can roll down to 1, this effectively means the 6th finger spell has a 'payload' of (Enchant - 2) (i.e. you can add phemes upto Enchant - 2 to this spell and be guaranteed to succeed).
Enhance pheme adds 2 difficulty and adds 1 Enchant. Therefore, by adding nothing but Enhance pheme to 6th finger, one can increase Enchanting like:
E(n+1) = E(n) + floor((E(n) - 2)/2) + 2 = floor(3*E(n)/2) + 1
written as a recurrence relation (the +2 is from the spell itself), where E(n) is the enchant skill after casting 6th finger n-times (at maximum power each time, of course).
So your enchanting skill increases exponentially with the number of casts. Assuming this is the case, this implies that there should be a fixed value of casts at the end of the sequence that you want to use the payload for non-enchanting related stuff (before that point, you get more benefit from increasing your enchanting skill). By inspection, an example:
E(n+1) ?? 2 * E(n) + 2 (this is comparing 1 cast at maximum power versus 2 casts at 2nd to highest power - the +2 is because the spell itself still provides 2 additional 'payload')
=> floor(3*E(n)/2) + 1 ?? 2*E(n) + 2
=> 0 ?? ceil(E(n)/2) + 1
=> 2 casts at the end is better than 1, always.
Since (3/2)^4 > 5 but (3/2)^3 < 4, we would expect the cut-off to be <5.
Example:
Suppose you have 10 Enchant skill and 7 time periods you can use to buff your stats. Then your skill goes like:
10 -> 16 -> 25 -> 38 -> 58 -> 88 -> 133 -> 200
If you'd used the last cast to buff, you'd have been able to spend 131 points on the buff phemes (2 to ensure your cast succeeds).
If you'd used the last two, 86 + 88 = 174
Last 3 = 56 + 58 + 60 = 174
Last 4 = 36 + 38 + 40 + 42 = 39 * 4 = 156
So optimal stop point would have been with about 2-3 casts remaining. If you'd just buffed stats from the beginning, you would have only gotten 8 + 10 + 12 + 14 + 16 + 18 + 20 = 14*7 = 98 points of buffing (and 'only' 20 points of enchanting) - about half (it would also be harder to fit in phemes with 7 different casts, since most phemes don't seem to cost 2). Then again, I guess the question is if you really need 174 points worth of phemes (I suspect for some things, the answer might actually be yes).
As a side note, this would seem terribly broken if this is how it actually works...the downside I suppose is that it does take a fair bit of time to set up, but the duration of 6th finger means you might be able to have absurd stats for at least a couple of days. Something else to note is that you could potentially be able to 'roll' the buffs - that is, on the last time period with your ridiculous enchanting level, you cast another 6th finger at maximum enchanting boost. This essentially means you can keep the high enchanting level as long as you cast the spell at least twice a week (which also means you can get the other boost whenever you need for effectively one cast).
Edit - apparently there's a maximum of 9 of a pheme per spell, but since Arid adds 2 for 4 difficulty, it seems like the maximum you can carry this out is to at least +54 (which is around 80 skill). After that the gains become linear, so there's diminishing returns on improving enchanting instead of the actual skill.
Edit 2 - Actually, Arid is in opposition to 6th, so doesn't work. Maximum practical boost then may be only about 22 (9+2, 2x), which also essentially gives base enchant x 2 + 4 'free' payload for maintaining the enchanting score (plus whatever buffing you do during the rest of the week).