As I understand it, anything above a very light wind you want a full sail and as wind builds you flatten to reduce increasing drag. Is boat speed the only way to tell if you have it right?
Aerodynamics doesnt support a full (draft) sail in between very light to moderately light wind speeds. The reason - not enough energy in the wind to keep it 'attached' as a boundary layer to the leeside - invisible separation stalls where the airstream becomes 'detached' from the sail. This can be predicted/assayed by the use of 'field effect plotters', or visually with tell-tales. Tell-tales (streaming correctly to validate that the boundary layer hasnt 'broken free') AND a speedometer are the principal means to verify the correct shape w/r the fore/aft position of the point of maximum draft along the cord of the sail ..... and its 'depth of draft' (full or flat).
The same applies to the very sensitive shape of the leading edge shape of the lee side luff portion - the luff 'entry' (flat or 'rounded') ... again tell-tales are the visual indicator to denote if or not the presence of separation stalls. If the airstream doesnt have sufficient energy to 'make the turn' around the front of the luff ... then separation stall.
Again at the leech exit portion, the wind must be 'leaving' the sail with the air stream velocity on both sides (on the lees side and the windward side) of the leech at EQUAL speed! - known as Kutta condition .... tell tales streaming straight back; if not, then the Kutta condition is not met and you have a separation stall. Tell-tales for visual verification plus the speedo.
Separation stalls (broken air stream boundary layer) are invisible - the sail doesnt 'shake' and the only way to verify/maximize is with a full set of tell tales. The principal areas of separation stalls are at any place where the airstreams 'make a turn' - leading edge, at the point of maximum draft, and the leech (kutta condition). You cant get optimum speed out of a sail if there are any invisible separation stalls present anywhere on the sail .... the only way to do that is by using a FULL set of tell-tales - tales at the luff, midcord, and leech.
Sooooooo, after you validate the absence of separation stalls .... and re-shape as necessary the sail by applying tension to an 'edge' (halyard, vang, traveller, cunningham, barberhauler, etc.) while watching the tell-tales to insure 'attached flow' ......... then when all 'tell-tales' are flying correctly, THEN you adjust the amount of draft (via outhaul) while watching the speedometer and the tell-tales AT near the point of maximum draft.
So, the answer is: The shape of the sail wont optimize until the Tell-tales are flying correctly ('shaping' the sail by inducing correct tension to the 'edges' of the sail) .... then set the correct/optimum amount of draft while watching the mid-cord **telltales** AND the **speedometer**.
The 'general rule' is FLAT sails for 'speed', FULL sails for 'power'.
--------
Some parenthetical 'asides'....
1. with boltroped dacron sails, the sailmaker intentionally cuts the boltrope shorter than the luff dimension (by approx. 1 inch for every 11 ft. of luff length) - called 'luff pre-load'. This so that the boltrope (and luff) stretches out to the full design condition when the sail is used at its 'designed maximum wind range' - usually 15-18 kts.
When you raise the mainsail, you usually need to 'stretch out' (by 'extra' halyard and/or cunningham tension) this 'pre-load' on the boltrope so that the fore/aft position of maximum draft is ~correct. If you dont add this extra tension to the luff, the point of max. draft will be too far aft in the sail .... and the boat will develop 'weather helm'.
2. What you learned in USA high schools, and what you see in most 'sailing books' about how sails 'work' is totally totally wrong, and has been since the Wright Brothers discovered that it was 'otherwise'.
Aerodynamics isnt 'intuitive', sails dont have 'thickness'. What makes a sail work is ....... the air streams have 'viscosity', and the friction from the viscosity of the air contacting the sail imparts energy to the sail ... and the net airflow 'circulates around' the sail/sails !!!!