D.A. Arnold "Optical and Infrared Transfer Function of the Lageos Retroreflector Array", 1978, NGR 09-015-002 (P179) confirms that they're for infrared measurements, and adds some interesting info:
The infrared array carried by the Lageos satellite was designed to provide coverage from any direction of illumination with a minimum of interference between the reflections from the different cube corners. Germanium was chose (sic) for the material because its high index of refraction gives each cube corner a very large viewing angle so that only a small number of reflectors is required. The material has the disadvantage that it become opaque at around 100C, which occurs if the cube corner facts the sun. An accurate range correction to the center of the satellite requires knowing the orientation of the satellite. In principle, the orientation can be determined from the infrared data.
It turns out that LAGEOS tracking is so incredibly accurate that having disparate reflectors caused some controversy over whether they were causing small anomalous offsets in the measured orbit. Two useful references:
C.f. D. Vokrouhlicky and G. Metris "LAGEOS asymmetric reflectivity and corner cube reflectors", JOURNAL OF GEOPHYSICAL RESEARCH, 2004 VOL. 109
followed by
D.M. Lucchesi "LAGEOS Satellites Germanium Cube-Corner-Retroreflectors and the Asymmetric Reflectivity Effect", Celestial Mechanics and Dynamical Astronomy March 2004, Volume 88, Issue 3, pp 269–291
The front surface reflectivity of the germanium CCRs is significantly higher than those of the fused silica CCRs; indeed, LAGEOS 1 photometry reported by Avizonis [1997] demonstrates sunlight reflections from germanium CCRs appear about twice as bright as those from fused silica CCRs.
and that in turn might be or not be (depending on the paper you read) the reason for the anomalies.
With $\Delta p$ ~ 0.015 and $\thera_r$ ~ 90°, both appropriate for the initial 20 years of LAGEOS (Figure 1), the asymmetric reflectivity is basically an acceleration along the spacecraft spin axis with magnitude of 12 pm/s2. It should be pointed out that the primary need to introduce the asymmetric reflectivity for LAGEOS orbits is to fit the anomalous along-track signal, where it contributes equally (or more) as the Yarkovsky - Schach effect during the periods when the satellite orbit enters the Earth’s shadow. Though not entirely negligible, the asymmetric reflectivity contribution to the eccentricity vector excitation is minor, about an order of magnitude smaller than the signal produced by the Yarkovsky-Schach effect and a slight net recalibration of the surface reflectivity [Metris et al., 1997].
Most LAGEOS tracking is done by the member stations of the International Laser Ranging Service. Most, perhaps all those stations measure with 532nm green lasers. Although many of them use 1064nm IR light to create the 532, I don't know of any that use the IR light for measurements.