You could open IW/Scheme, use IW to gain lots of maps and scheme to play the IW as much as possible. By not buying non cantrip cards, you'd keep your effective deck size small and quickly gain a large TM density mitigating the luck factor of the card. Then use your newly gained golds to grab those provinces, picking up silver with the IW once you're done with going for maps.
I COULD do that... but not doing that did get my 4 Provinces in 13 Turns. Do you think your IW/Scheme/TM plan could really do that well? I obviously got great shuffle luck, but let's say it had taken me 15 Turns to get those 4 Provinces. Worth it?
My first suggestion was just off the top of my head. I went and did a little solitaire and well, it's kinda hard to know how well it needs to work. I found that with the IW/Scheme opening, you often get $2 for an embargo. Which would obviously be played on the gold pile if the opponent is not going for maps thus denying them that card. So with these embargoes, a benchmark is kinda hard to set. So it really is hard to compare this to your strategy. Could you have gotten your provinces when you did if the gold pile was embargoed say around turn 7 or 8ish?
However, in the solitaire games, it did suffer from the luck of the draw. Sometimes I would have 4 maps in my deck after the second reshuffle, sometimes only 2. Sometimes they hit, sometimes they wouldn't. So this strategy is very subject to luck of the draw, but it also has some ability to hinder the opponents. It probably isn't the optimal strategy, but it has the potential to be great.
So to wrap up, here's a good ole chemistry analogy. Treasure map games are like
Mössbauer spectroscopy. When they're good, they're really really good. However they are only really really good in certain games.
*For those interested, Mössbauer spectroscopy requires the source to emit gamma rays. It is therefore limited to certain isotopes of certain elements and thus can only be used in limited cases. But when it is used, the spectrum it provides has an extremely high sensitivity.