After actually reading the yeast section in "How to Brew" it's reinforced my personal preference to allow the starter to ferment and for the yeast to settle before pitching.Depending on the style/gravity etc I sometimes pitch yeast (slurry) but in many cases I want an active starter. It's easy to adjust a recipe to allow for additional unhopped wort, plus a little sucrose to readjust fermentability.
JP suggests that if the stater wort is "very similar" to the beer-wort then pitching an 'active starter' is most likely the best way to go. This is because the yeast in the starter has produced a specific set of enzymes for the wort's sugar profile, so it's ready to go straight away.
However, if the composition of the stater and beer-wort are different (especially true if using a starter containing refined sugar), by pitching an 'active' starter it may actually impair the yeast and fermentation.
By letting the yeast ferment and settle they build up glycogen and trehalose reserves, allowing the yeast to start with a 'clean slate' and so it is better prepared to adapt to the new wort, resulting in less lag time and better fermentation than if it was pitched when the starter was 'active'.
When creating a starter I either use LDME or wort saved from a previous brew, so the starter wort will always be different to the (AG) beer-wort, as a result I feel there is more reason to let it ferment fully and settle before pitching. Only pitching the yeast-slurry also reduces any off-flavours that might have been produced in the starter.
But yeast are hardy little buggers and this is not really rocket-science so there are many ways to do it and equally many ways to produce good beer, which is what we all aim to do, no matter (if or) how we create and pitch a starter.