Try this
run a mile carrying 2 bags of shopping
then run it again without them
shall we bet on which is the fastest time?
Well if its Dessie doing the running, and if he's required to do it straight afterwards, I reckon he'd be faster the first time.
It doesn't work like that. There reaches a point where a horse won't go any faster regardless of how low a weight you give it. We are talking about strong and powerful animals.
I did make the mistake (alright I did it deliberately) of mentioning it in a pub once to Oxford physics graduates (they loved applying their intellect to risque things like speed rating and horses - thought it gave them a bit of edgy credibility on the street I reckon). It's called hysterisis or 'chatter' in an enginerring context.
Basically a horse is a strong animla and can carry weight to a point where it isn't affected to any point of note. Weight reductions and performance improvements do not make for a perfect linear regression if you plotted them. A curve would be concave in its nature. If it were a straight linear regression then horses would run at silly speeds.
They are afterall limited by their cubic capacity, like a combustion engine is. Unfortunately it's a Mordin quote I draw on, but I've never seen a better way of illustrating it.
You've got a car with a 1 litre engine (he used the example of his Nissan Micra - draw your own conclusions) and say it weighs a 2000Ibs laden with passengers etc and has a topspeed of 70mph under these conditions. You then take all your passengers out and strip it bare of anything that adds ballast to it. Hypothetically you strip it down to 500Ibs, effectively reducing the mass by 3 fold through weight reduction. It wouldn't improve on its topspeed and suddenly run at 210 mph.
Apologies if the maths is wrong, (drinking some
American wine) but the principle is the same