When this was being modelled (in a medium, medium/heavy and heavy) we decided on a through rotation to ensure consistency of results. As Horror Box has said, the actual pitch angle will vary with loading and conditions (of wind). Under design conditions the wind conditions (usually light to calm) and C of G has to provide a deterministic set of conditions for flight test. (I doubt if any manufacturer has tested a take-off procedure under all of the conditions found offshore.)
However, this is slightly off thread because Crab was asking a question about 'flyaway'; the only reason I introduced the results of the modelling was because it was from experience and reasonably hard numbers could be quoted.
It is well to remember that, for a take-off, we usually are describing a 'dynamic' manoeuvre where there is already energy in the manoeuvre before the failure. For a flyaway manoeuvre the start point is static with respect to vertical energy (although there might be some wind effect).
If the flyaway is being conducted from somewhere near OEI Vstayup the loss of height will only be the result from the remaining engine ramping up to max power (a hydro-mechanical will involve lag because the rotor slowing will provide the cue).
As someone else has mentioned, beneficial effect of rotation will only apply when the mass is low enough to enable a recovery; under all circumstances, an aircraft will ditch if the mass is too high for recovery. Under condition where ditching is the inevitable result of the engine failure, rotation is not required to effect a OEI recovery but to minimise the rate of descent.
To rotate to 20 degrees when ditching is inevitable, might place the aircraft in a high rate of descent and accelerating condition.
It may be a dumb thing to say but the best condition to be in is knowing: what will be the result of an engine failure; and, if flyaway is possible, what the Vstayup and Vtoss is and (as asked by Crab) what the optimum rotation will be for recovery.
As previously described, a small rotation with light wind conditions will only work if the mass equates to HEC D mass; conversely, a large rotation with a higher wind will (unnecessarily) sacrifice drop down.
Lastly, it is not just the rotation that has to be considered, it is also the time that the pitch angle is held. With some manufacturers, it is recommended to ease back on the cyclic as soon as the angle has been achieved; with others there is a Vt (target speed - before Vtoss) so as not to overshoot the Vtoss.
Jim