More specifically, faster twists are needed as bullet LENGTH increases, or as velocity decreases.  A .22-250 can use a slower twist rate than a .223 with the same length bullet, because the velocity is higher, which gets the bullet spinning faster.

LOTS of things impact accuracy, and twist rate (meaning: bullet stabilization) is just one of them.

The original XM16s had a 1:14 twist, as this was the common twist rate used in the .222Rem bolt guns at the time.  However, the .222 guns were used for varminting, and used short, light bullets.  The 1:14 twist wasn't always able to properly stabilze the 55gr FMJ bullet, especially in arctic conditions, so it was tightened to 1:12 for the M16A1.

The 1:7 was a compromise between the twist rate ideal for long M855/SS-109 bullets (1:9) and the VERY long M856 tracer bullets (1:5).

If we're talking about leadcore bullets, 1:9 will properly stabilize anything from 40gr-75gr.  The 1:8 twist is used to stabilize the long 80gr bullets for long-range competition.  These are the longest/heaviest bullets in general use.  1:7 is really too fast unless you're using M856 tracer on a regular basis.

-Troy

On the first issue, it is very common for folks using light varmint bullets (which have very thin jackets) to blow apart in 1:7 twists.  For example, the 40gr Speer TNTs.

About the second: despite the long-repeated myth, the twist rate of a rifle has NOTHING to do with the damage caused by the bullet (assuming the bullet is stabilized enough to hit point-first, if not, wounds will be LESS severe).  VELOCITY, not twist rate, is the reason why M193 and M855 ammunition fragments, causing such large wounds.  But the bullet must strike the target at 2700fps or greater to ensure fragmentation; otherwise, the round is far less effective.  A shorter barrel will impart less velocity, and reduce the range at which fragmentation occurs.

-Troy