Dom... I'm glad you went some way towards taking my advice I did say don't go faster than 9:15 per mile, but I'll not quibble
This answer is a bit long... but I don't have time to edit it to smarten it up... but read carefully, and you'll get my point.
I'll stick to my original advice. Stick to slow runs for most of your runs in this fortnight. There is a reason. In simple terms, it is best to build a base fitness before moving on to speed work. This involves laying down systems for blood supply to muscles.. systems that are specialised in transferring oxygen in and waste products away... good aerobic systems. Once you've got the systems in place (extra blood capilliary etc),, you can tune them to be more efficient as you increase your speed, which has particular impact on the removal of lactate (and other) waste products.
If you train "too fast" (7 miles in 57mins)... your body is under pressure and has to use the network of capilliaries and cells that are already in place. It won't really add to them but will tune them a bit to make them a bit more efficient... so you will see some improvement.
Much much better if you increase the size of network before thinking of improving its efficiency. But you're in this ridiculous position of wanting to do it all in 4 weeks! I really don't know the perfect advice. At one extreme, you could concentrate solely on slow running and building up the network... and maybe you end up with decent endurance, but not the speed for the task. On the other hand, you might do all your training as fast as possible... and get some great speed... but then conk out in the race after 3K.
You need to find a compromise... and my suggestion was an attempt to do a bit of both. It's accepted that your body takes 2-3 weeks to adapt to "slow" training... which is one reason I'm saying don't waste your precious time on fast stuff yet... because you only have a limited window of opportunity to build that base network of capilliaries etc. Focus now on slow training speeds that do not recruit your 'fast' systems. Then, in the 3rd week have a go at improving the efficiency of your newly-built networks.
Other people would perhaps offer a different compromise solution, and they may or may not be right... but I'd just warn that intuitively, when faced with training, we tend to think faster is better... but, for the reasons outlined, it can be far far better to be patient and intelligent with your training.
I'm trying to think of an analogy (and failing to get a realistic one)... but imagine you had a battered 1000cc car and knew you had a race in a week's time - but you're not sure of the quality of the opposition. Maybe you could choose to spend all your time tuning every component of that 1000cc engine to be as fast as possible. But you don't know if you're up against 2,000cc engines... in which case, it doesn't matter how brilliantly you fine-tune the small engine, you're going to lose (and also by screwing everything out of your small engine, there's more chance of a breakdown!). Maybe it's better to invest time in installing a bigger engine before thinking of fine-tuning. Not perfect.. but there's something in that analogy - and highlights the dilemma. In my opinion, you should invest at least some time on increasing the size of your engine before trying to make it fast.
This was in response to a question about improving very quickly, but the "sc