Has the development of double consonants in Latin been studied?

When one studies both Latin and Greek, one of the most prominent differences between the two is the much greater number of double consonants in Latin. While Greek does have some instances of them, they are everywhere in Latin.

While the origin of some of their occurrences is straightforward to recognize (f.e. prefixes, ad+locare > allocare), others are not. They can appear in core words of the language, which aswe know are generally of native and not borrowed origin.

Cognate words in Greek (and other Indo-European languages) often lack the double consonants.

Examples:
ferre, quattuor, vacca, pessimus, mittere, caballus

I was intrigued by this, as Indo-European itself did not have geminates and there is often no obvious reason for them to be geminates, what is most visible with quattuor (an obvious reason would be assimilation of a nearby phoneme, e.g. Latin septem > Italian sette). Thus, they evidently appeared in the language at some point. However, I have failed to find an explanation for this development, despite having consulted some specialized literature.

Has this issue been studied and the rules behind the occurrence of geminates in Latin been established by any scholar? As my research was not so extensive and Latin has been the object of intense study, I feel the need to ask.