When prefixing a vuv with a sh'va, if the next letter is a sh'va as well, it will put two sh'vas in a row, which is not possible at the beginning of a syllable. So, the vuv with a sh'va is changed into a shuruk.
A shuruk is a t'nuah g'dola, meaning it is a full syllable, thus the following sh'va must be na, opening the next syllable. Note, a sh'va always opens or closes a syllable. To open a syllable it must be the first letter. To close it must be the last letter (or, according to some, not be followed by anything other than a another sh'va.) Since the shuruk already closed the previous syllable, the sh'va must open a new one.
In our case, a sh'va has become a shuruq. We can question, do we see it as a full fledged shuruk or do we see it as a sh'va masquerading as a shuruq? In the former, the the shuruk is a t'nua g'dola and the sh'va is na, as above. In the latter, the shuruk is a t'nua k'tana and the sh'va is nuch.
Even if the shuruk is a t'nua k'tana, a meseg changes it. Because it is not a "real" t'nua k'tana, stressing the syllable transforms it into a full shuruk, making it a t'nua gedola.
I do not know who the originals are. My understanding is the GR"A and Rav Yaakov Emdin are of the latter opinion.