The Picture Splinters discussion in your book begins with the classic 2/1 GF >auctions, and with trump already agreed, and we have little trouble with those. Add in the minor complication of using the jump to set trump, and we still seem OK.
However, when we get to the auctions which aren't already GF we reach dangerously different interpretations. To help us sort this out, which of the following (and why) are reasonably Picture Splinters versus being ordinary splinters, and which might not be splinters at all. I've tried to have enough variety to let you clarify the definition, or maybe you'll just be able to state it so obviously you won't need to torture yourself with all these examples.
1. West East
Here, I would expect 4C to be a Picture Splinter, with a high likelihood of Qx(x) in spades. Why? I can cuebid in response to 2S as a game-acceptance on most hands that would qualify for a club Splinter that are not right for a Picture Splinter (better spade honor, lesser trumps, lesser diamonds).
2. West East
Here, I would not expect a Picture Splinter, without discussion, because there is no second suit shown by the person splintering. I could see describing this as a fit-showing Picture Splinter (diamonds being the assumed second suit -- great double fit), especially as Responder has many options with hands that include a stiff spade.
3. West East
This seems identical to (2.) -- same analysis.
4. West East
Here, I would not view this as appropriate for a Picture Splinter, because there is no second suit held by Opener. Further, there is no assumed suit possibility (provided by Responder). So, I would expect this to be a "normal" splinter. However, I would strongly suspect 5440 (a void splinter).
5. West East
No Splinter at all, the way I play. I'd expect Responder to have 5-5 pattern.
6. West East
This one is weird, as I use 4th-suit GF. The jump also takes away RKCB, and, if a Splinter, the minor fit is unknown (unless a default is agreed). If I had to guess at the table, one thing. If I was allowed to discuss this and settle on an agreement for this specific auction, I'd prefer Redwood (4D as RKCB for clubs, 4H as RKCB for diamonds).
7. West East
1C 2C (inverted minor)
Similar to above; best use may be a Picture Splinter, with the assumed second suit being diamonds. Note how this principle (not discussed in the book) might be viewed as similar to the Picture Jump to 3NT after a major is agreed. Note also how useful this might be in practice. "If I have shown only a raise, then my Picture Splinter shows good trumps, good support in YOUR second suit (two of the top three honors), a stiff, and no control in the fourth suit."
8. West East
Normal Splinter. No second suit can be inferred in any way.
9. West East
Same analysis as for 1D-1S-2C-3H.
10. West East
If discussed, this seems like another good candidate for the principle of a Picture Splinter, with the assumed suit being spades (probably KQx+).
11. West East
1S 3C (invitational)
I'd expect a normal splinter, because of limited options.
12. East West
Same as (11.). The problem with having these be Picture Splinters is the inability to clearly define alternative auctions as cuebidding.
13. East West
Same as (11.) and (12.). An additional point. After simple preference auctions, into a minor, the Moysian fit possibility in Responder's major adds one more level of options and confuses the issue too much, IMO.
14. West East
Clearly a splinter. IMO, not Picture Splinter because of option limitations. I dislike the idea of making a "convenient" but false jump shift into a major just to establish a force so that I can later raise a minor, all to ensure purity of a Picture Splinter in support of the minor. Too much for me. I want 2H and 2S protected as legit. So, normal Splinter for me.