There are two textual variants in this passage, one dealing with the order of the words never and hath, the other with the choice between hath and has. The second change (made by Joseph Smith in his editing for the 1837 edition) is, of course, secondary; the critical text will restore the original hath here (for a complete discussion, see under infl al endings in volume 3).
In the original manuscript, Oliver Cowdery originally wrote “hath never”; but then he crossed out the hath and wrote it inline after the never (which shows that the correction was an immediate one). Joseph Smith, in his editing of the printer’s manuscript, crossed out the hath after the never and inserted has supralinearly before the never. Thus Joseph apparently intended to change the order so that the have verb would precede the never. But in the setting of the 1837 edition, the earlier word order (in 𝓟 and in the 1830 edition)—namely, with never first—was maintained.
Elsewhere there are considerably more examples where never precedes the perfective have verb instead of following it: 25 times versus 7 times. In one other case, there has been some word order variation in the manuscripts, namely in Alma 26:22 (see the discussion there). So in each case, we follow the early textual sources in determining the position of never with respect to the perfect auxiliary have. Here in Alma 30:28, we follow the corrected reading in 𝓞: “a being which never hath been seen nor known”.
Summary: Maintain in Alma 30:28 the placement of never before the perfect auxiliary verb: “a being which never hath been seen nor known”; also restore the original form for the perfective have (namely, hath).