That sounds pretty much like what I do. 10 minutes into the boil I pull a 2L flask for my starter. At this point the hot break has all ready occured and the wort remains unhopped. I chill the rest of my wort with an immersion chiller, flush the carboy with CO2 before I siphon the wort in, then flush the headspace with CO2 after. I pitch the 2L starter into my wort 24 hours later. Since I pitch the whole starter, I only use O2 at the beginning of the starter, I don't swirl or stirplate so I don't oxidize it, and I grow the starter at the same fermentation temps to avoid excessive esters and alcohols.
Since I pitch the whole starter I want it to be the same grain bill as the wort. I don't like the idea of crash cooling the starter and then decanting for a couple of reasons. I don't think you can really crash out all the yeast in 24 or 48 hours, definetly not the lower flocculating strains, like a kolsch. And I don't like getting the yeast all ready to go at fermentation temps, then dropping the temp down to make them dormant again, only having to raise the temp right back up again to pitch. These temperature fluctuations have got to stress the yeast at least a little bit.
We all got our own ways of making a starter, and in the end I think most of us are just trying to make the best beer we possibly can.
Definetly hit the wort with O2 again before you pitch the starter in.
