The Filter^: Price discrimination

I'm revising course material and want to do a better job of teaching price discrimination. Economicshelp.org outlines the Pigouvian triumverate:

1. First Degree Price Discrimination
charging consumers the maximum price that they are willing to pay

2. Second Degree Price Discrimination
charging different prices depending upon the quantity consumed.

3. Third Degree Price Discrimination
charging different prices to different groups of people

I don't like this for various reasons, and was interested in the Wikipedia section on Ivan Png:

Complete discrimination -- where each user purchases up to the point where the user's marginal benefit equals the marginal cost of the item;
Direct segmentation -- where the seller can condition price on some attribute (like age or gender) that directly segments the buyers;
Indirect segmentation -- where the seller relies on some proxy (eg, package size, usage quantity, coupon) to structure a choice that indirectly segments the buyers.

Robert Frank's classic micro textbook:

First-degree: charge separate prices along the MR curve
Second-degree: Price declines with the quantity you buy
Hurdle model: discounted price available to who jump a hurdle

According to Tim Harford in The Undercover Economist:

First degree/Unique target: evaluate each customer as an individual and charge according to willingness to pay
Second approach/group target: different prices to members of distinct groups
the third way/self-incrimination: persuade customers to reveal that they're not sensitive to price

Tutor4U is virtually incomprehensible, demonstrating the need to put things together in a clear framework. Ideally, I think the following should be made clear:

Firms would ideally treat us as individuals (first degree/unique target/complete discrimination) and whilst we see evidence of this (real estate, antiques) it's extremely costly and rare. In fact, if we define it as most textbooks do it's impossible in the real world. Therefore there should be a distinction between the firm's aim and their strategy to carry it out. Hence one "method" of price discrimination is "individual bargaining".
Since this is costly (not just because firms don't "know" our demand curves, but because "we" like price clarity and certainty) there are strategies that attempt to use proxies for individual bargaining
- Self-identification - either through coupons or slight product differentiation
- Group segmentation - based on semi-aggregated demand curves

The reason why Harford's self-incrimination is not the same thing as Frank's hurdling is that in the former the product is different. True, to all intents and purposes the costs are the same but this denies subjectivism and risks being wrong. Similarly quantity discounts is not the same thing as group segmentation because 100 packets of photocopier paper is not the same thing as 10 shipments of 10 packets.

Here's what I propose as a synthesis. Comments welcome.