I had what I think was the first 4400 sold (serial 10046) - trialled over Christmas 1990 and took formal delivery in March 1991. I added a second one a couple of years later. They were not co-axial units, rather heavy but extremely stable in any conditions. They could be set up all day ina howling gale and at the end of the day the RO wold be within a few seconds - very important when you are robotics.
The very first targets were a dummy instrument with just a built-in tilt sensor. radio and keyboard - lightweight but bulky although they were perfectly balanced for carrying. It mounted on a prism pole and in use you plumbed the pole, aimed the sight at the instrument and pressed the button. The tilt sensor sent the VA to the instrument which then searched at that elevation, so finding the target was very quick. It was so good you could shoot through heavy foliage and, once, even through the hinge gap of a gate.
By the time I had the second instrument Geodimeter had gone over to the block targets, but I continued using the tilt target whenever there was heavy foliage or other obstructions. I looked at the 600 series at the time. It had the advantage of being co-axial and much lighter, but the range was considerably less. In trials the 4400 consistently gave between 10 and 20% better range - significant when in poor conditions.
Eventually, around 2007 time there was some big project out in the far east and Geodimeter (Trimble) bought up just about all the available instruments in the UK to ship them out east. Very good trade-ins were offered so I swapped over to a 640 and 5605.
Incidentally the 4400 series was preceded by the GDM464 which was a test-bed for the 4400 technology. It was a fully servo driven GDM400 using a class 2 laser. Only a few were built and were sold into controlled environments, mainly mining and engineering, where the safety of the laser could be ensured. I did some field trials on a prototype, mainly building elevations where it performed much like a modern day reflectorless.
I also did one geotechnical trial using three of them (at that time only four had been built) on a ground collapse test. This involved piling up concrete blocks, each of 2 tons, onto a test slab of concrete road with a whole series of markers around the area. The 464s were used since they would drive straight to each target with a single keystroke, leaving the surveyors to just make the final pointing. Speed was important when each observation sequence needed to be as close as simultaneous as possible - 48 targets if I remember correctly.
The test took around 48 hours and finished when the pile reached 6 blocks high and 750 tons, the ground turned to jelly under the heap which collapsed. The top blocks bounced everywhere - although the ground "jellied" under steady load it stayed hard on impact loads, so instead of sinking in the falling blocks rebounded. Just as well we had the instruments set back by about 50 metres. The post test survey was interesting, locating all the ridges, shapes and hollows once the blocks had been lifted clear.