If the outside temperature – on all sides of the room – is 0 degrees and one heater (which is much hotter than 50 degrees) will be able to reach 25 degrees in the room, two heaters will reach 50 degrees. That's true for the temperature change over a particular timeframe; it is also true for the equilibrium temperature (the temperature we reach after some time when the leaking heat cancels the heat flowing from the heating device).
The reason is that all the heat fluxes are (almost exactly) proportional to the temperature differences and the heat fluxes are equal to the extra heat coming from the heating device which is also doubled.
In reality, the final or equilibrium temperature will be lower than 50 degrees because of some of the reasons below or their combination:
A part of the warmth in the room isn't due to the heating system but due to surrounding apartments or other spaces whose temperature is higher than 0 degrees. One may say that the average outside temperature is higher than 0 degrees. If it is $T_a$, then two heating devices will achieve $T_a+2\times(25-T_a) = 50-T_a$ which will be lower than 50 degrees if $T_a$ is greater than zero. Alternatively, we may say that the heating is just a part of the temperature increase from 0 degrees, so doubling the heating makes a smaller impact.
The heating itself isn't much warmer than 50 degrees. For example, if the heating's own temperature is 40 degrees, then 40 degrees is also the maximum temperature one may achieve in the room even if we combine many heating devices of the same kind. The temperature will only converge to 40 degrees from below, so the addition of the temperature differences will be sublinear.
When it gets much warmer than 25 degrees, someone will find it really annoying and will try to turn the heating off.
Some insulators around the room may have heat conductivity that (usually slighty) increases with the temperature, so the heat losses at higher temperature will be greater than expected from the proportionality to the temperature difference.