Uniface, that is a very solid/comprehensive answer, and a lot more concise than I can be... And it's an answer that also draws out a very neat distinction between Kirk cluster and St. Charles/Plevena cluster points (I'm going back to saying Dovetail because it's easier to type.)
Kirk sites are common in the Midwest from Ontario on down, but especially common in the Ohio River valley. Some of them like Swans Landing, and the Cesaer's Casino in Indiana site near Louisville have produced hundreds and hundreds of kirk points all made with very local materials, as well as animal bones, tools, lithic workshops, tube bannerstones, etc. even some burials. The animal bones show that Kirk peoples ate a lot of rabbit, racoon, turtles, fish, with far lesser numbers of what we think of as traditional game animals like deer, elk, bear. They probably didn't hunt over huge areas, and utilized local resources. Dovetails are found over a similar area, but to date very few dovetail sites have been found and none even remotely as 'rich' in information as even a minor kirk rockshelter. It's really not clear who the Dovetail people were, but as Uniface pointed out, we know that during the early archaic period they controlled several main quarries (Flintridge, Harrison/Wyandotte County hornstone, some of the Coshocton and other black flint quarries in Ohio, etc.) All of which lends credence to the idea that Dovetails were an imported/traded commodity, while Kirks were made locally with local materials. Note, there are some dovetails made with local materials, which could have been transference of technology, but most were probably made at centralized locations. As you get to Illinois/Missouri they might have been made on a more local level, explaining the greater variety of St. Charles/Dovetail blades found in those states.
Back to the idea of Cobbs being preforms for archaic points... Personally, I think too many knives are called Cobbs blades. I think there were true preforms (either thinned quarry blanks, or large core bifaces that were tools which were reduced clovis style down to another tool, and late stage bifaces that were going to be made into a specific point) and then there are a series of hand held, unnotched knives that include the Cobbs that were finished tools. Cobbs have some unique traits and in my opinion were a hand held knife that was resharped and reused to the point of getting down to a drill like form and then discarded. Their distribution is heavier south of the classic range of doves, and nearly absent in the northern part of the range. Kirk sites produce a lot of smaller Cobb like preforms, but so have some sites that produce Lost Lake and Thebes points as well. They haven't been found on Dovetail sites, but then again there are only two or three small dovetails sites so the sample is probably too small for conculsions.
If any of you get a chance to see Bob Converse's collection, or pictures/pieces from Stan Copeland's collection, there are un-notched dovetails (big flint ridge bifaces built like dovetails) but I don't think they are related to the Cobbs type. I think most classic ohio-style Dovetails were notched when made, and then progressively resharped after trading. (Heck, look at all the examples of 5, 6 and the few 7 inch finished dovetails with very little resharpening.) Even two of the pieces posted in this thread are large pieces with heavy percussion flaking, and very little retouch. I don't think ancient peoples would use Cobbs blade as a hand held knife, resharpen and reuse it; then decide to reflake the entire surface with percussion flaking, and finally notch it to get a finished Dovetail.