The odds of this happening at Google in the next ten years are pretty low, I'd say. That's why they organized this as a members-only union rather than trying to do a real unionization drive.
Members only does lower the bar considerably.
Size matters. If it gets big?
New ideas matter too.
The traditional union struggle has become very difficult in the US. The high solidarity numbers required today are 20 to 30 points more than necessary before.
Solid negotiations basically pay the dues and then some.
This may not be true for all diciplines / positions.
In terms of risks, marginalizing those is worth something too.
For many, just keeping health care sane is worth a lot more than a couple points.
There is no such thing as "improving things for everyone" because different people have different preferences. Some people want job stability, some people want work life balance, some people just want to get paid, fuck everything else. If a union tries to pursue some priorities over others they are screwing over all the employees who have different preferences.
And frankly, sometimes there is no basis for solidarity, and thus no union.
Resolving that is a discussion that actually does determine whether there is improvement for everyone, not just some blanket statement or other.
Finally, yeah. A few people may not give a fuck. Consideration due is consideration given.
If your union refuses to let workers be exploited for unpaid overtime, but you really like working to make other people rich, leave!?
In the US, costs and risks are pretty high. People need more than they did in times past.
What terms improve things for everyone? How should work be?
From there, they build solidarity around those terms.
When the vision maps to 80 / 90 percent of labor at that enterprise, the effort to unionize can win.