Nonviolent methods often require the possibility of violence as a backdrop to be effective. Otherwise, they might not yield the desired results.
While Gandhi's philosophy sounds nice in theory, it may not always be the most practical in real-world scenarios.
In her book Civil Resistance: What Everyone Needs to Know:
* https://www.goodreads.com/book/show/44096650-civil-resistanc...
* https://www.goodreads.com/book/show/10056014-why-civil-resis...
Erica Chenoweth has an appendix listing six hundred movements dating from 1900 which are classified in how violent they were. She found that those that used violence (more) succeeded in achieving their goals 25% time, but those that did not use violence (at all, or much less) succeeded over 40% of the time: you almost double your odds by eschewing violence.
Further, movement that were violent and succeeded were more likely to be oppressive/authoritarian (possibly because the movement leaders internalized the possibility that the same methods would be used against them: the overthrown often don't end up in pleasant places in those situations), while non-violent ones were less likely to be (though no guarantee, with 1970s Iran being the main outlier).
So it appears that the general historical record seems to support Gandhi's philosophy.