Here is a something worth knowing about, from Pires, D.P., H. Oliveira, L.D. Melo, S. Sillankorva, and J. Azeredo. 2016. Bacteriophage-encoded depolymerases: their diversity and biotechnological applications. Appl. Microbiol. Biotechnol. 100:2141-2151. [PubMed], (calls to figure and table excluded from quote):
Based on our search, the huge majority of phage depolymerases (126 proteins) are encoded in the same open reading frame of phage structural proteins (mostly on tail fibers, base plates, but sometimes also in the neck) or in close proximity to those genes, and were thus considered as structural proteins. Twenty other depolymerases found in this work might be soluble proteins since they are distant from any structural gene.
Depolymerases that are only soluble, that is, not virion attached, presumably are only useful in the immediate vicinity of phage-lysed bacteria, e.g., towards phage burrowing more deeply into biofilms. This perhaps means that phages don’t need depolymerases to initially infect biofilm bacteria (see here for that argument). Depolymerases that are associated with virions, by contrast, presumably are useful as well upon initial phage encounter with a biofilm bacterium.
That the majority of depolymerases are may not be soluble, but instead appear to be associated with virions, is suggestive that depolymerases are employed for the sake of initial encounter between virions and biofilm bacteteria. But this then begs the question of why more phages don’t encoded depolymerases?
Is it that we have trouble recognizing them in sequence data? Is it that bacteria are just too diverse in terms of extracellular polymers produced? (In addition to limiting utility, the latter may also interfere with our ability to detect depolymerase phenotypes during phage growth as plaques.) Is it because for the most part phages can infect biofilm bacteria sufficiently even without depolymerases? Or are there unexplored trade-offs associated with depolymerase encoding, perhaps especially when they are present as structural components of phage virions?
In the three previous paragraphs I am drawing on a tiny bit of past thought as to the role of depolymerases in phage interaction with biofilms, as can be found in my 2011 book, Bacteriophages and Biofilms. In particular, from p. 23 (of the revised pagination version, or p. 27 of the original… don’t ask…):
Ecologically, EPS depolymerases improve phage movement that occurs either adjacent to or distant from a phage’s parental infection. If distant, then movement towards bacteria will be enhanced by physical linkage between virions and depolymerases. Alternatively, for more localized movement, then soluble depolymerases may suffice, such as for phage dissemination out of biofilms . Scholl et al.  thus found that efficiency of plating (EOP) was low for phages encoding a soluble EPS depolymerase when infecting a K1 capsule-producing strain and that an isogenic phage not encoding the depolymerase is “unable to form plaques on lawns of this strain” (p. 4872). This result is suggestive that though initiation of plaques occurred with low efficiency, once those infections commenced then subsequent EPS depolymerization presumably facilitated phage migration towards adjacent bacteria to complete plaque formation. In circumstances where enzymes may not be directly supplied, it should thus be advantageous for those enzymes to be carried by virion particles, if only to increase the efficiency of initial infection. That is, it should be advantageous to phages for enzymes to be present at the point of phage adsorption, by being virion attached, rather than present only immediately following the lyses of phage-infected bacteria [I then illustrate this argument with a figure…].