I make the mistake sometimes of not "carrying over" the physics of one device in the ST universe onto another.
You can hardly be faulted for that; indeed, much Trek fiction is written where the author goes overboard with "connectivity" and the result is a needlessly small and cramped fictional universe. "Phase technology" just happens to be a hobby-horse of mine, sort of, as it's 100% fictional in all its supposed applications but the terminology is alluringly consistent.
Are they devices which measure the information state of the traveller, destroy the original and recreates him elsewhere, or are they, as James Blish once wrote, devices which causes the traveller's particles to make a "Dirac jump" to the destination?
There is some onscreen dialogue to support the idea that the transporter scans the target in detail and stores the results as pure data, but it would seem to be trumped by the very explicit concept of the transportee then becoming a swirling mass of something they call a "pattern", stored in a "pattern buffer" until sent forward as a "phased matter stream". It's not an abstraction you can store in a hard drive and extract at your leisure, it's a physical state of existence, even if an alien one.
How is abstract different from physical here? Well, the main difference would seem to be that there always only exists a given quantity of this "phased matter", exactly corresponding to the transportee: you can't copy it. Pure data should be eminently copyable in perfect detail without major effort, while physical matter intuitively should not.
Lots of ambiguity there. But what's clear is that both the transporter and the phaser somehow "cheat" in that they definitely aren't managing the sort of energies involved in turning mass m into energy E=mcc and back. If mass is merely turned into phased mass using rules of physics we don't know about yet, the technology becomes more plausible overall. That is, one big cheat covers a lot of bases and eliminates the need for multiple other cheats in explaining how transporters and phasers could work. Anything
ought to be better than E=mcc...