I am wondering if I have broken my sensor, or if someone else has seen this problem and perhaps found a solution?
At least it worked around 4 years ago, but only you can tell how it has been treated since then.
US sensors are piezos (only cut for a higher frequency than your run-of-the-mill buzzer) and the natural enemies of piezos are DC voltages, high temperatures and physical shocks, which all can depolarize them, or even cause breaks in the crystal in case of DC, if the voltage is high enough. A partially depolarised TX will send out a weaker ping and a RX will receive less from a given amount of reflected signal.
Partial depolarizing will cause a response as you experience, as the max range depends on the signal being sent is powerfull enough to reach an object that reflects a sufficient amount to be picked up by the receiver and without known good parts, it's hard to tell if it's the TX, the RX or both that is weakened.
While I cannot be 100% sure if that is the cause in your case, your measurements support it - and if it is, there's no way to "recharge" them, but in case it's just stiffness or accumulated dust and grime, try running it for a day or two, by just letting the controller ping it continually, as fast as possible - if you measure the return signal at start and after an hour or two and it hasn't improved the received signal strength a bit, it 's probably a waste of time to continue though.
I have no experience with the Devantech US units, so I cannot say if it may be a software issue, or if it's even possible that it can be.
Searching google for "weak range srf05" has produced nothing.
Maybe try other search terms - strictly speaking, "weak" when talking distance is nonsense - "weak output" or "short range" might yield a better result.