Personally, I prefer not to be too technical but if I may say so...
Relationship between stride length and stride frequency is, just as you have pointed out with other topic, not that simple as it has been presented; if you're not long-strider, then you're "choppy" strider. Everybody has the correct combination of two that best suit to him/her. I don't think it's something you can "perfect" long-stride running then quick-strude running. Besides, if you "perfect" one and to the "perfection", why move on to "perfect" the other end of spectrum? Curiousity? Research?
I used to think, like you said, stride length is better/easier way to improve. Thus hill bounding exercise, etc. But, as you know, it's not that simple at all. Bounding or springing exercise Lydiard advocates WILL enhance stride length which is most likely underdeveloped among distance runners who run a lot. However, this should not be attained with the expense of stride frequency; otherwise, all you're doing is to waste extra energy by moving the center of gravity up and down. Besides, long-stride approach actually has more obvious limitations.
Also another consideration is that, as the distance goes longer and the conservation of energy becomes more critical, "long-stride running" could potentially become disadvantage--particularly if you have a tendency of "bounding", you are more likely to use your calf muscles instead of thigh muscle groups, which are much bigger and endure better. In general, in the marathon, long-stride runners tend to run very fast if he/she hits a good one but more likely to hit the wall--not because he/she runs out of gas but because their "kicking" muscles get tired. Choppier stride runners tend to run more consistantly though may not run as fast or respond to the surges better.
Sprinters DO have difinate differences among them whether they are long-stride type of short-stride type. Long stride type, like Carl Lewis, can accelerate a lot but have difficulty starting quickly.
Definition of "Speed Endurance", as much as I understand, is pretty much close to Lactate Threshold...or something like that. The speed at which you consider "speed endurance" training would have to be the speed at which your body start to create lactic acid or slightly above or blow and develop tolerence to continue running at that speed for a long time; which, as you have described, is probably close to one's 800m pace. However, according to your definition, which was "how long you can maintain a given speed", no specifics of the "given speed" is stated. In such case, the man who runs ultra marathon, because he is maintaining the "given speed" for a loooooooooooong time; thus has the best "speed endurance" which I'm sure you'd say is wrong (sorry, I'm being too technical and smart-Alec here).
My understanding of "speed endurance" is somewhere between "speed" and "stamina". Whether it has to be above "lactic acid formation" or below that, I don't know and there are quite a few who lurk this thread more qualified to explain terminology better than I can. Of course, if you want to compete in 800m, you probably need to work on "tolerence" to lactate...whatever you want to call it. If you are training for the marathon, you probably want to develop (make it higher) the pace at which you can maintain without creating lactic acid in your working muscles.
I'm not quite sure if I understood your comment about 800m correctly. Speed does play a big part in running 800m and, if your "basic speed" is slow, all the anaerobic ability or lactate threshold or "speed endurance" or whatever will get you to the finish line ahead of the faster guy. To me, 800m is different from "speed", and doesn't have much value or indication to good performance for running 5000, 10000 or marathon. It would help if you have better time. But why stay there and try to improve when your best suited distance is somewhere else?