Some interesting points of view.
I think it's a combination of the points mentioned and the track which throws the results.
It's a leaders track with a short straight, but also a fairly testing turn into the home straight. The camber of that turn is quite tight by comparison. Basically you have to be on the pace or leading, and able to handle a hard turn and hang on in the straight.

Therefore, although the strike rate is fairly consistent with other tracks, the horses generally start over the odds next start.

You could add Geelong to that list

Bunbury Races today:
Chocolate Moon and Angel Wing are last start winners at Bunbury. Both are early favourites. Does this fit the criteria ? Will be watching them - even though I usually don't bet on WA races...

Chocolate Moon just won - sprinting away from the field! Paid at least $4

Angel Wing finished 4th. However, with a $1 bet on both horses, you would have made $4. That's 200% profit...

The winner of race 8 was also a selection I believe, except that it was favourite.
So, two from 3 with one official selection which won at 4.0 local tote.

Its ok if its favourite - with the "revised" second lot of rules. That means 2 winners out of 3 that were last start winners at Bunbury!
That's $6.30 from a $3 total bet. Sounds good to me...

Looking back to 26 March (last Bunbury Races), there were 2 picks for a 3rd and a win at $4.40.

1/1/2014 to 31/3/2014

Last start winner at Bunbury
Racing within 30 days
Non metro track this start.

47 selections
11 winners
23.40% strike rate
Profit $1.20 (TAB price)
POT 2.55%

Place Profit $11.40 (TAB price)
24.26% POT

Does the $93 winner distort the profit result?How likely is it that a huge outsider will win again?

Of course kiwi, but looking at the prices of winners, longshots often bob up.
They are sent out way over their true odds.