The back-ups 350, CS500, and similar small offline UPSes have 3.5W idle consumption, measured with fully charged battery and a calibrated, verified to perform correctly, effective power meter*.
Back-UPS Pro 1500, line interactive, is 6.3W measured with the same meter.
I briefly measured a SmartUPS RT3000 (two-man lift even without batteries), and I believe it would have gone below 50W, but I did not have time to let the batteries charge.
Here is a back-of-envelope tip regarding power dissipation: Consider the outside surface of the UPS as a heat sink, and roughly measure the area of the sides and back. Also add the top, but since the top is not vertical, only add half of that. For the BackUPS 1500, that area comes to about 300000 mm^2. Now, from much experience, I know that a vertical 64000 mm^2 one-sided surface will rise 1K when you dissipate 1W in it. Since the UPS is 4.7 times larger, 6.3W will cause the outside to rise 1.3K. If I average the temperature over the case using my IR thermometer, and subtract the temperature of the carpet it is sitting on, that is about what I get.
Your 700 is smaller, with an outside area of about 100000 mm^2. This means it takes 1.6W per K temperature rise, and with 50W it should be 31K above room temperature. My carpet is currently 27C, so it would be 58C average case temperature, and hotter in places, assuming the fan isn’t going.
*This meter works by sampling tension and current at 50kHz, and then averages the products over a number of periods with first a sliding window and then a low pass filter. The point is that I know exactly how it works, and I can verify it.