Let's calculate the required acceleration for you 100 years in 3 years scenario, fortunately Wikipedia provides the relevant formula:
when you start with
g is the constant acceleration. But you don't accelerate the whole time; upon your return you want to stop. And at half the travel distance, you probably have to stop and turn around, unless you fly in circles (of ever-increasing and then-decreasing radii due to the centrifugal force).
So, assuming the expression above holds symmetrically for deceleration, the most likely trip would be far away, stop and return, i.e. accelerate 0.75 years, decelerate 0.75 years to far away, and return the same way for 1.5 years, and each of those four phases should last 25 years on earth. That yields a required acceleration of 78.2 m/s², or about 8g. Potentially survivable, but for 3 years this sounds unpleasant.
Let's go in circles instead, so we only need to accelerate and decelerate once for 1.5 years each, both taking 50 earth-years. Then you only need 39.1 m/s², or about 4g, which is apparently something mere mortals can basically take.
So your time numbers aren't that unrealistic, the problem is of course to find an actual spaceship that can do this for 3 years nonstop, plus you should carefully plot your course (since you'd travel up to 4.6 ly in that time).
The real problem is propulsion: If you go by conventional rockets (or rather, a relativistic rocket) and assuming the best possible exhaust velocity, namely
c and that you arrive back at Earth with mostly no rocket left (
m_1=100 kg), you need a rocket with an initial weight of 1.8e359 kg, which slightly exceeds the mass of the observable universe... The circular travel vastly improves this (4.3e108 kg), but not remotely enough.
Summary: While the acceleration required is bearable, achieving it via conventional means of propulsion is impossible. You'd probably have to be very creative with thousands of Swing-by's (which would also influence the passage of time) in order to get along with realistic amounts of fuel, and at those high momenta you would probably severely influence the assisting celestial bodies. Sounds like a lot of havoc to merely travel into the future...
note I think I mixed up external and internal acceleration in the G-Force determination, the perceived force might actually be larger...