No, there is nothing that is "keeping light from going faster". The local velocity of light in vacuum can not be different than the standard $c=3 \times 10^8{ m\over{sec}}$. There are two parts to my answer.
1) When light passes near you in vacuum, you will always measure the standard $c=3 \times 10^8{ m\over{sec}}$ using your local meter stick and clock.
However, if you are far from a star (a gravitational potential well) and the light passes in vacuum near the star, you will see, using your local meter stick and clock, the light going slower than the standard c (see https://en.wikipedia.org/wiki/Shapiro_delay for experimental proof). In fact if the star were a black hole and the light was near the Schwarzschild radius, the light's velocity would approach zero.
Conversely, if you were near the star and the light passed far away, you would see, using your local meter stick and clock, the light passing at greater than the standard c. In fact if the star were a black hole and you were near the Schwarzschild radius, the far away light's velocity would approach infinity.
So it is possible to view the stuff called light going at different velocities if the light is not local.
2) The notion of being able to change the LOCAL speed of light in vacuum arises from a misconception of what c is. Our historical way of measuring velocity (and the speed of the stuff called light) in meters per second, is based on comparisons to a meter stick and a second stored in the Bureau of Standards. But this was before Special Relativity in 1905 and the non-abelian boosts of the Lorentz Group. We now measure velocity in radians of Lorentz Boost $\lambda$. We can sense how much $\lambda$ a boosted body has by watching a clock riding on the body. The clock will appear $\gamma=cosh(\lambda)$ times slower than an un-boosted clock. Inverting this gives $\lambda = cosh^{-1}(\gamma)$ radians of boost (the modern velocity) for the object. We can relate radians of velocity back to the historical stick per tick velocities by ${v \over c}=tanh(\lambda)$ but this is not necessary to do modern physics. By letting v=c in this formula we see that the stuff called light has a modern velocity of $\lambda=\infty $ radians. These modern velocities add when pointed in the same direction. Historical velocities do not simply add, and you must use for them a special addition formula involving c.
The constant c and other velocities based on a standard stick per tick in the Bureau of Standards are historical artifacts. After 1905, c and v no longer need to appear in physics. Now, objects have dimensionless velocities in units of radians. If your original question referred to changing the constant c, the question is irrelevant because c is not needed in modern physics, and anyway c just depends on what stick and tick are stored in the Bureau of Standards. If your original question referred to changing the velocity of the stuff called light, then it is from changing from $\lambda =\infty$ to something else like what? $\lambda = .8 \times \infty ...or..maybe... \lambda=100+\infty$ ?? ... all of which are still $\infty$.