I think that comparing cells (and organisms in general) to a 'machine' is mostly a result of humanity, as of yet, lacking a better, more precise go-to analogy for 'the thing that a cell is like', much in the same way we often compare the human brain to a 'computer' even in spite of realizing that this is an insufficient descriptor of what a human brain actually is.
The issue, of course, is that when we start using this sort of shorthand terminology, it's bound to get people to start taking it over-literally.
This is one of the problems I have with the more recent trend of equating organisms with algorithms. It's not that the analogy isn't apt -- in fact it's startlingly effective (for those still startled by such things) -- but rather I feel it risks repeating the same reductionist perspective that leads to the issue at hand.
Barring any forthcoming breakthroughs in the appropriate scientific fields, a cell isn't anything other than a cell, and a cell -- like the human brain -- is something we just do not yet fully grasp the nature of.