Great game David - and a very different experience to the Defense games.

However, I think (for me) GeoSpark would benefit from the following gameplay tweaks to make it really stellar:

1/Scoring:

The game is relatively unstructured (in terms of levels). That means that the only goal is for a high score. As such, there needs to be a real drive within the gameplay for getting high scores. Moreover, to make it addictive, there should be increased risk associated with those super high scores.

At the moment it would (is) actually very easy to keep tacking up scores simply by tapping each enemy individually (10 points each). There is almost no risk associated with this, and the high score purely based on time spent playing = boring.

Obviously, there is an increased score for chains (which are riskier). However, I think the score increase for longer chains should be greater - thereby increasing the highest scores possible and effectively penalising the one-tap approach, and really testing the player's nerve.

Current score for each enemy in a chain is: 10, 20, 30, 40, 50, 60, 70, 80, etc

Two alternative ideas for score patterns are

1: double the points for each enemy in a chain - i.e. 1, 2, 4, 8, 16, etc. Thus a chain of 8 would score (1+2+4+8+16+32+64+128=255). And super long chains of 20 could score very high: =1048575 (2^10 - 1). The benefit of this model is that getting an additional enemy on the chain would be worth as much on its own as the entire value of the chain up to that point - a real worthwhile risk!

2: However, the doubling model might be too extreme. Instead, how about a triangular number pattern with the score for each enemy in a chain being: 10, 30, 60, 100, 150, 210, 280, 360, 450, 550 etc. With a chain of 8 = 1200.

3: Actually, a Fibonacci sequence might be a much better compromise: 1, 2, 3, 5, 8, 13, 21, 34, 55, 89, 144, 233, 377, etc. Short chains pay lower than the triangular model, and longer chains more, but not as much as the doubling model.

Importantly, the generation of new enemy types (i.e increased difficulty) should be based on the total number of enemies destroyed, NOT on time or total score. Thus, there is a reward in trying to make chains (high scores) as fast as possible before it gets more difficult. 

Conversely, bonuses (x2) and extra lives, should be based on score, NOT on time or enemies destroyed. Therefore you get more bonuses the faster you accrue points.

Total enemies destroyed (in addition to total scores) could also me reported at the end of the game. Most skillful = highest score from fewest enemies. Perhaps the score could be divided by the enemy count to get a skill rating?

Together these points help to force the player to make chains (if they want anything approaching a high score / high skill level)

2/ Other gameplay ideas:

Differentiate the enemies. For example, make the reds into a special bonus multiplier. i.e. Each member of the red chain might not score any points itself, but could increase a bonus multiplier. This multiplier would be applied onto the next enemy or enemy chain destroyed. i.e. make a chain of 8 reds = 8x multiplier. Make a chain of 6 pinks: get the sum of this pink chain x8. Multiplier then resets to x1. Multiplier could work in addition to bonus x2 multipliers. This would add an element of strategy in selecting which enemies to eliminate first.

Multitouch. Make it possible to use another finger to eliminate individual enemies (by tapping them) whilst still holding onto a long chain. Could open the door to extremely long chains - with ever-increasing gravity eventually making it impossible to hold on without death.

That's it for now! Sorry or the long post. Keep up the good work. Very entertaining.

Cheers, Matt