Okay, the practicalities!!!
Firstly- need to run a 4:20 mile- yes and no. Firstly you need a lot of speed and a 4:20 mile may well be a pre-requisite, howver...assuming there is a pre-requisite "time" (not saying what it is) if that time was accomplished off a very high lactate tolerance and anaerobic system ( similiar to that of a top class 800m runner), you'd need to go faster as basically, a lot of the components that make you a great 800m runner (i.e. anaerobic ability), compare very unfavourably with the abilities of a longer distance runner who runs that time off of a strong high vo2 max, who would need a lesser time. Though the factors of stride length + efficiency still come into play.
Basically, Paula Radcliffe can run say a 4:20 mile, and because of the way she's made, she'll be able to convert that very well. Kelly Holmes can run a lot faster than 4:20, but couldn't convert that time at all, as everything that lets her run a very fast 800/mile, doesn't convert well to longer distances.
Obvious stuff, but it needs to be taken into consideration! My 800 PB is 2:11 (2:08 in a TT) whilst my training partners is 1:59, yet he in now way has the pre-requisite time that he'd be able to make a good or even reasonable 10 mile runner ( to give you an idea he has a sub 2 800 and his best 1500 is over 4:30 (only just under 5 m/m)). On the other hand, I'd like to think that if I got my 800 time down to 2:01/2, I would have the prerequisite. That's assuming I didn't accomplish it through soley working my anaerobic ability!!
Tom- I sense a lot of Frank Horwill in that 10 x 1 mile- stripping it to the basics and then going at it.
I think it sounds like an excellent attempt to see how you can pace yourself and to determine just how much work you need- I know I could run sub 5 m/m pace for 10 x 1 miles, the question is how much rest would I need ( certain on this as I've run more than 10 4:30 1500ms!). At the moment I'd predict around 5 minutes recovery per rep needed.
However- whilst this would give me an idea and to be hoenst, its such a hard session it'd be very useful, I think that training scientifically, you could accomplish your goal quicker training in a different way.
Namely by improving these 3 aspects of your running...
1) vo2 max
2) LT
3) running economy
and there are certain ways to train each of these systems.
Alex makes a veyr good point when he talks about controlled intensity, the aim in training shouldn't be to run at 100%, it should be to run at 90% but controlled.
1) 3k-5k pace is that which improves your VO2 max hte most, so training at this pace for reps of 2-5 minutes, with half recovery times of work times with a total distance of 5-8k is a very handy device. This gets you used to running fast at pace as well.
2)LT training is also going to be extremely useful, and perhaps more so in this case! A session like 10 x 1 mile would really work it hard. Training at about the pace you could run at for an hour.
3) economy- apparently- there are three ways I've heard of that are meant to be effective in improving this
i) hard speedwork- mile or so pace, plenty of recovery, improves footstrike, leg strength etc.
ii) flexbility/ drills
iii) just plain running for a lot of years! improvements in mitochondria and the like help improve your running economy.
OKay, lot of credit I think going to mr Daniels. here!
Think a useful thing we could try coming up with on here, is a list of pre-requisites, and a plan for how to get someone to break 50 minutes for 10 miles. Be quite interesting I think!