...
Whenever an already stated POSSIBLY_EQUIVALENT_TO target itself also become inactive - whether at the same release or later - the combinatorial logic of associations should be:
A POSSIBLY_EQUIVALENT_TO (B OR C) and B SAME_AS D implies A POSSIBLY_EQUIVALENT_TO (C OR D)
A POSSIBLY_EQUIVALENT_TO (B OR C) and B REPLACED_BY D implies A POSSIBLY_EQUIVALENT_TO (C OR D)
A POSSIBLY_EQUIVALENT_TO (B OR C) and B MOVED_TO D implies A POSSIBLY_EQUIVALENT_TO (C)
A POSSIBLY_EQUIVALENT_TO (B OR C) and E D MOVED_FROM B implies A POSSIBLY_EQUIVALENT_TO (C OR EOR D)
A POSSIBLY_EQUIVALENT_TO (B OR C) and B POSSIBLY_EQUIVALENT_TO (D OR E) implies A POSSIBLY_EQUIVALENT_TO (C OR D OR E)
A POSSIBLY_EQUIVALENT_TO (B OR C) and B WAS_A (D AND E) implies A POSSIBLY_EQUIVALENT_TO C, WAS_A (D OR E) * tricky one - suggest detailed review!