There are nearly as many ways to evaluate hand distribution as there are experts. In some cases, these methods are chosen to justify being passive or aggressive in certain circumstances. However everyone recognizes two distinct components to evaluating distribution: a) the core value of the distribution; and b) the fit with partner.

In the days of Goren, a relatively simple evaluation system, one evaluated short suits only, for both suit and Notrump contracts, as:

 - 3 for void;
 - 2 for singleton; and
 - 1 for doubleton.

However remember that Goren's NT openings were all one point stronger than today, and all experts deduct a full point for 4333 distribution, When we shift to modern NT ranges and properly downgrade 4333, we end up with something like this:

 - +3 for void;
 - +2 for singleton;
 - +1 for doubleton; and
 - -1 for 4333.

Now 4432 distribution  with ***15-17 HCP*** with exactly matches Goren's 16-18 Total Points; but we have a more accurate and precise evaluation for other distributions. 

Every expert I know also counts one point for the fifth card in a suit, even if it is expressed as a shift in the opening NT range: ie "*Don't open 1NT with a five card suit and 17 HCP.*" or "*I open 1NT with 15 HCP and a five card suit.*"

We now have:

 - +3 for void;
 - +2 for singleton;
 - +1 for doubleton;
 - +1 for a fifth card in a suit; and
 - -1 for 4333

For additional cards in a suit, these are (at least usually) worth more than the single point for a fifth card - but how much more is unsure. Sheinwold recommended in **Five Weeks to Winning Bridge** to count every card past the fifth at two points each, but that may be a bit aggressive. it may be better to regard it as having value both intrinsically and due to fit (or lack of it) with partner. With this in mind, we couldsay something like this:

 - +3 for void;
 - +2 for singleton;
 - +1 for doubleton;
 - -1 for 4333;
 - +1 for a fifth card in a suit; and
 - +2 for a sixth card in a suit when partner shows at least two; and
 - +1 for a sixth card in a suit otherwise; and
 - +2 for seventh and subsequent card in the suit.

Back to your specific example, this evaluates 5332 distribution as (+1+1) - (-1) = +3 points better than a 4333 hand; but only 1 point better than a 4432 distribution.

As a final note, Goren recommended not bidding 3NT until the partnership was known to have 26 total points. This is regarded as overly cautious by modern standards. Players are better overall than they were two generations ago, and as we see above modern evaluation is more refined. The modern standard is 25 points for Matchpoints, Rubber, and Not Vulnerable at IMPS; and 24 points Vulnerable at IMPS. As I have discussed in other answers, a partnership should agree that the one point variation for vulnerability at IMPS (and the complementary adjustment for slam bidding) is made ***only*** by the hand inviting game, or bidding it without an invitation. This is the only agreement for it I am aware of that is both unambiguous and simple to describe and remember.