Seems to me that this is a pretty slight variation. Probably normal variance in components. I dont own either one, and dont feel the need to look up their accuracy claims on the internet, so Im going to make some assumptions here.
The computers only display 3 digits, right? Lets pretend for a moment that we could extend the displays by one decimal point, so that they now display XXX.X feet instead of XXX feet. Assuming normal rounding off, the other computers could have been reading anywhere from 99.5 to 100.4. Yours could have been reading anywhere from 98.5 to 99.4.
Now lets go back to the 3-digit displays. Worst case scenario, your computer is measuring 98.5 (displaying 99) and theirs are measuring 100.4 (displaying 100), the difference is 1.9 feet. Best case scenario, yours is measuring 99.4 (again, displaying 99) and theirs are measuring 99.5 (again displaying 100), a difference of 0.1 feet.
What I'm trying to say is that the actual difference could be as little as 0.1 foot to give a 1 foot displayed difference. Even in the worst case scenario, assuming that the other computers were spot-on (absolutely no error), your computer had less than a 2 foot error. A 1.9 foot difference at a depth of 100 feet is a 1.9% error. At your safety stop depth (where the reading is more critical) a 1.9% error would be about 3.5 inches. Pretty darned good.
Without a more precise display (more digits displayed), there is no way to know exactly what the computer difference was, and either way, the readings are well within the design accuracy of the computers. My opinion is that you're just not going to find a computer or gauge that accurate in a price thats generally affordable, and it would be overkill for diving. Besides, it would be a pain to use, since the readings would fluctuate every time a large wave passed overhead.
Between the four computers, you could have had actual readings of 98.5, 99.0, 100.0, and 100.4 feet. and still get the readings you reported with three digit displays.
In other words, dont be fooled into thinking that yours was wrong just because it was the odd man out.
WRT the other part of your question, I don't know the sampling rate of the computers, so don't know how long you would have to hold your arm still for them all to 'normalize.' But, again, I wouldn't worry about such slight differences.