Changing atmospheric conditions may influence a storm's extended forecast path. Forecasting a storm days before impact is like predicting a vehicle's movement on a highway days in advance. You may have a sense of that car's destination, but it has multiple paths to reach its destination. Depending on what traffic, construction or otherway roadway hazards it hits, the vehicle may take a detour.
Even on a small scale, think about how many different routes you could take to work. What determines the path you take? Any number of factors could re-route you.
Like a road, the atmospheric highway has steering currents that influence a storm's path. Those steering currents may come in the form of a high pressure, a low pressure, a frontal boundary or other atmospheric conditions. For the purpose of this explanation, I'll simply call them 'steering currents.'
If you change the time the steering current (i.e. the vehicle) gets on the road - or the road in which the steering current takes - you could cause a change to the overall forecast. Even something as small as 'changing lanes' could change a forecast.
I admit the highway metaphor is an oversimplification of complex hurricane forecasting. I hope it help explains why the path has moved - and likely will again.
I understand no explanation will help burden the stress of an approaching storm. There are always a lot of questions: How bad will it be? When will it get here? Do I need to leave?
Emergency management officials have the impossible tasks of deciding whether or not to issue evacuations. Due to the time it takes to evacuate masses of people, such an order has to be issued days in advance, given when the storm forecast margin of error may still be high.
Ultimately Hurricane Irma is going to take the path of least residence. The storm's lifespan will end when it becomes too weak to 'defend' itself from steering currents and ultimately breaks apart or combines with other atmospheric conditions.