Preliminary analyses of COBE/FIRAS data have already produced a spectacularly accurate determination of the microwave background spectrum for 1 cm ≤ λ ≤ 500 μm. The absence of detectable deviations from a blackbody spectrum sets strong constraints on physical conditions of the intergalactic plasma and, in particular, has ruled out the possibihty of a truly diffuse thermal bremsstrahlung origin of the X-ray background. General arguments suggest that comptonization distortions due to heating of the intergalactic medium associated with the formation of cosmic structures, with hot protogalactic winds, or with the ionizing flux from AGNs, are likely to be very small (comptonization parameter y ≲, 10-4). A larger signal is expected from the integrated re-radiation from dust in external galaxies; to what extent this may conceal possible comptonization distortions depends on the maximum redshift at which galaxies contain substantial amounts of dust and on the temperature distribution of dust grains. In any case, a precise determination of either the y parameter or the background from distant galaxies requires a careful subtraction of the emission from the Milky Way.
The great success of COBE strengthens the need for a parallel improvement in the accuracy of spectral measurements in the Rayleigh-Jeans region, where imprints of physical processes occurring at very early epochs (such as, e.g., the dissipation of small scale density inhomogeneities) may show up.