Probably the only realistic thing that Blizzard could do is have those at Bronze, Silver and Gold Leagues have a somewhat randomize set of start-up conditions so that players can’t literally play out a recipe strategy they read on the net. But I don’t see that happening.
I think you're on the right track here, but I disagree on the specifics. I do agree that higher starting resources would be more interesting, and randomizing those resources might
work (I'd be cautious of Zerg's larva production bottleneck causing them to relatively overperform the other factions at some resource values and underperform at others) but I don't think it would at all stop the "all-in" gambits. It would make their timing and composition a bit more variable, but people would still be pulling them off and quitting when they fail.
My belief is that there are two issues at play. The first is that all Starcraft II 1v1 maps are extremely
similar, adhering religiously to the formula of a high-ground starting position with a single narrow entrance (possibly a secondary entrance blocked by rocks) and secondary resources right outside of this entrance. They tend to be similarly sized, the distance between bases is meticulously measured out, and the intervening terrain is usually fairly straightforward and clear. With such reliably homogenous starting scenarios, players can be confident the same defensive and offensive tricks will ALWAYS work the same way, which is boring and defeats the entire point of different maps. Unless the game reaches the late game, all 1v1 maps play essentially the same way, and this means opening gambits always
work the same way.
The second, and much more important issue, is the very high lethality and decisiveness of Starcraft battles. They're over in seconds, there is virtually never an opportunity to retreat, and there is no way to regroup even if you could. This means that it's often easy to win decisive battles, and virtually impossible to recover from a decisive loss. Unless you're perfectly
evenly matched, this usually means the first battle has an obvious winner and he's going to have to screw up to lose the match. This is a remarkable departure from Warcraft III, where you actually could buy a "get out of jail free" card for 350 gold and had very effective "natural" defenses in your base to give you a chance to regroup, so a single bad field battle was hardly a game-ender.
It's very sad that Blizzard abandoned everything they learned in WC3 and were afraid to at all tweak the formula of Starcraft, creating a game that essentially has the same strengths and weaknesses of its predecessor.
Personally, if a game like Starcraft 2 came from a smaller studio, I'd be praising it. But coming from Blizzard with virtually limitless resources and development time, I'm actually quite underwhelmed. This is the very best that a world-class developer with no timing pressures and infinite development resources can produce? It's a great game, to be certain, but in my opinion it's inferior to Warcraft III.
EDIT: Try playing large team (3v3/4v4). The maps are quite a bit more variable, and you will rarely run into coordinated 3 or 4 man gambits. Without coordination, a solo gambit is very easy to foil so you usually see conservative openers (note: conservative != passive; a standard ling, zealot, marine, or roach rush is conservative, but it can still be very, very deadly)