Well, first, the vast majority of wells use fracing, and always have.
Second, its not 2 or 3 thousand wells. There's better than 12,000 working wells in the ANF alone, and over a hundred thousand in the state of Pennsylvania, the vast majority of which were frac'd at least once in their histories. If you add in the ones that are no longer producing, and the # of ones that were frac'd more than once, then you probably have over half a million frac'ings in our state that were already carried out.
If you go just by recent history, the renewed interest in gas in PA happened around 2007, 2008. The # of new wells PER YEAR went from just under 1000 to closer to 1500 for just the ANF, which is a county sized area. The numbers outside of the forest, for a similar sized area, are similar.
Now, I've seen exactly 1 case in PA where a well water contamination was linked to gas drilling. That was in Dimock. And that affected between 10-20 houses. The company was heavily fined, it was a fractured well-head. There was also a spill of frac fluid in SW PA, a pipe burst. Sad indeed, and these incidents should be investigated and prevented. But look at the percentages.
Marcellus is new, and there are some Marcellus wells already producing. The difference between them and the old wells is depth, and because of the depth, they drill horizontally at depth in several directions instead of making several separate vertical wells. Environmentally, this means that you'll have much fewer total wells, but the amount of bore space in a single well is much more. It needs a lot more water for the initial fracing operation, but like the other wells once that is complete, there is no need for continuous water. Due to the heavier drilling equipment, and more water you'll need, it means that the space to be cleared for the operation is larger, which means more siltation. It also means that due to the depth, the frac'ing operation is LESS likely to contaminate water wells. Of course, its still possible, and if the contamination comes from broken concrete above the water table like these other ones, it really doesn't make a difference how deep the well is.
So yes, well water contamination is a problem but not the big one. The big one is land use, water use, and water disposal, and thats where we, as conservationists, need to focus on.
Now, the story in Colorado:
First, its hard to compare drilling in Colorado to here. Different geology, different type of rocks, different elevation differences between valley and mountain. There, if you drill half a mile deep from the high ground, thats still above the water table in the valleys!
Weld County: 1395 producing wells as of the time of the article. 29 residents asked for testing of their water. 9 tested positive for methane that was verified to be natural, i.e. not related to the drilling. Exactly 1 was shown to be due to drilling activity.
Garfield County: This is a little more serious. 1300 producing wells in the county. "Dozens" of wells were found to be at least partially "degraded" due to drilling, though it doesn't state the extent. At least 1 case basically made a house unlivable. One single bad well casing can do that to "dozens" of water wells.