100% is not a good goal. Google's guidance treats about 60% as acceptable, 75% as commendable and 90% as exemplary, but says there is no universal ideal number and that chasing the last few points gives diminishing returns. Coverage only shows that lines ran, not that they were asserted correctly, and the value is mostly in showing what is not covered. Better practices are high coverage on changed code (per-commit goals of 90% or more are reasonable), reviewing uncovered lines in code review, and using mutation testing to check test strength. Fowler also advises against testing trivial getters and setters just to hit a number.

Sources: Google Testing Blog, Fowler