The comparison of the properties of galaxy cluster samples selected using observations in different wavebands may shed light on potential biases of the way in which the samples are assembled. For this comparison, we introduce a new observable that does not require previous knowledge of the cluster mass: the X-ray mean surface brightness within the central 300 kpc. We found that clusters with low surface brightness, defined as those with a mean surface brightness below 43.35 erg s−1 Mpc−2, are about one quarter of the whole cluster population in a sample of 32 clusters in the nearby Universe, selected independently of the intracluster medium properties. Almost no example of a low central surface brightness cluster exists instead in two X-ray selected samples, one sample based on XMM-Newton XXL-100 survey data and the other on full-depth eROSITA eFEDS data, although these clusters are known to exist in the range of redshift and mass as probed by these two surveys. Furthermore, the Sunayev–Zeldovich Atacama Cosmology Telescope cluster survey is even more selective than the previous two samples because it does not even include clusters with intermediate surface brightness, which are instead present in X-ray selected samples that explore the same volume of the Universe. Finally, a measure of the mean surface brightness, which is obtained without knowledge of the mass, proves to be effective in narrowing the number of clusters to be followed-up because it recognizes those with a low gas fraction or with a low X-ray luminosity for their mass. Identifying these would otherwise require knowledge of the mass for all clusters.