I think the simplest explanation for dark matter is that it consists of something we already know exists and behaves exactly like dark matter: black holes. To fit this explanation, we would need a vast number of black holes, each relatively small, approximately the size of a tennis ball. These black holes would be large enough to have persisted since the early universe, avoiding conversion into Hawking radiation. The…
Dark matter is non-collisional, doesn't decay, and doesn't strongly scatter radiation. The last is about gravitational microlensing. You address the first two in another comment, but focused on falling into central black holes of galaxies and Hawking radiation. Black holes merge (cold dark matter doesn't clump, and certainly the absorption cross-section of CDM does not grow with clumping in the event there is some "dark chemistry" in the CDM sector; black holes absorb more radiation as their masses increase, including via merger). The MACHO (massive compact halo object) models which use small black holes have to keep them very sparse and on a restricted set of orbits in order to keep them from merging in galaxy and galaxy-cluster halos where they would be detectable. It's also hard to have galaxy-cluster distributions of small black holes: the density profiles get very different as tiny black holes interact with each other. Additionally, in distributions around the central-bright-galaxy of clusters, it's hard to suppress inverse compton scattering X-rays from dust scattering around (or forming accretion structures around) tiny black holes. So the signature radiation a tiny black hole MACHO model has to worry about is less Hawking than the behaviour of normal matter moving around black-hole-infused space in clusters like Abell 2764 https://www.esa.int/ESA_Multimedia/Images/2024/05/Euclid_s_n...> or Abell 2744 https://esahubble.org/images/heic1506c/> https://en.wikipedia.org/wiki/Abell_2744> for example. It's really hard for simulators to arrive at this kind of structure using mainly MACHOs, and there seem to be lots of these structures. Lots and lots.
It's somewhat easier to do with ~solar mass and bigger black holes (they can be sparser, their orbits are thus less contrived to avoid collisions -- after all star-star collisions are rare unlike dust-dust collisions), although larger-mass BHs run into limits form gravitational microlensing. In particular, BHs of 100 or more solar masses would tend to disrupt wide stellar binaries, particularly in at the outer margins of galaxies (including our own, easy enough to look for out of the plane of the disc). And there are strong limits on BHs between about 10 and 2000 solar masses from ultra-faint dwarf galaxies, where domination at that mass of BH would expand the visible part of those galaxies (especially bright star clusters like the central one in the Eridanus II dwarf satellite https://en.wikipedia.org/wiki/Eridanus_II>) via dynamical heating.
The MACHO idea was certainly plausible, and projects like OGLE and MOA developed microlensing-observation techniques which are now used for finding low-mass and wide-orbit exoplanets, and "orphan" Neptune-like planets ejected from their birthplace star systems. See for example https://kmtnet.kasi.re.kr/~ulens/
However, there seems to be very little hope that black holes are a large fraction of whatever the detailed description of dark matter might be.
That there are plausible mechanisms for forming primordial black holes at arbitrary masses does not mean they are actually there.