On what basis can you make those claims (equal force, force-time, recruitment patterns, etc)? In sprinting the foot would contact in front of the hips and the hips would pass over indicating MUCH more use of posterior chain than in a high knee action. EMG studies support this. Ground contacts are also MUCH faster. And if mimicking was the goal, why would you use DBs in the hands rather than a weight vest?
The part of sprinting where this might apply is somewhere around the 5th step where the posterior chain is not the main contributor. Force-contact times would be anywhere from 130ms to 100ms and the use of DBs would change body lean dependent on where you hold them. Take a picture, draw the force angle acting through the hip, tilt the picture slightly to match the 5th step (or thereabouts) and see that it is pretty close. From that picture you can then change some of the biomechanics to better match sprinting. The main reason why you would do this is to be able to replicate the 5th step (or thereabouts) for enough repetitions to gain a learning effect. The types of learning effects that you may want to enable include preactivation and learning to reduce GCT whilst maintaining force magnitude; in fact you want to increase force magnitude over time. Using some simple physics equations you can calculate how much weight you need to lift (inckuding bodyweight) andthe height that you need to rise each step to math the force requirements of the 5th step (or thereabouts) in sprinting. If facilities allow, using a slight hill and marking each step may be a more accurate way to meet these requirements.
You can use a weight vest if you want…if you have a slope then you may not need to change the upper nody angle much.