While there's no problem connecting the power rails in parallel something you should be aware of with Sharp distance measurement sensors is that while the average current draw of 30mA shown in the datasheet is correct during the measurement cycle there are peaks of around 100mA. Here are some current measurements I just made of a single sensor at 1000 samples per second so the X axis is in millseconds:
I've found them to be very sensitive to supply voltage fluctuations and in your case you may have short peaks up to 500mA that your power supply may not cope with well. As well as 0.1uF decoupling capacitors across each device you may find it benefical to also place a larger value capacitor near the devices to help smooth out the peak current demands. Here is the latter measurement taken again with a 2200uF capacitor across the supply rails:
You can see the peak current has dropped to 36mA and the span of current measurements has dropped from 93 to 27mA which your power supply should be a lot happier with when you multiply those numbers by five. For the 0.1uF capacitors because the plastic enclosure extends somewhat below the PCB you can solder those directly to the sensor PCB without taking up any additional room.
Determining the best amount of capacitance will be somewhat experimental but if you find your readings are erractic try a single 2200uF for the entire sensor group as a starting point and see how things go from there, you'll probably find a smaller value will be fine. I've also seen suggestions such as here that adding a small amount of capacitance to the analog output can be beneficial if you continue to get noisy output once the power supply is clean.