In all, or perhaps almost all, places of religious worship, there is a need for the words spoken by the celebrant (which we use as a generic term for priest, rabbi etc.) to be clearly audible and intelligible to the congregation (once again used as a generic term to cover worshippers of any and all religions). The necessity for this is two-fold. Firstly, the celebrant is the leader of the worship. If his or her words cannot be heard, then the congregation is unable to take part in worship as they would wish to. Secondly, religious services often include an element of teaching. Obviously, if the congregation cannot hear clearly, they cannot be taught.
Most major religions pre-date the technology of sound engineering, as do many places of worship. How did people hear what was said in the old days? The answer is that they didn't. The unamplified voice can be heard in a large auditorium, accommodating up to around 2000 people in ideal conditions. But the acoustic design of places of worship often leads to a massive excess of reverberation, which clouds and colors every word that is said
If the voice of the celebrant is amplified, then this just makes the reverberation louder, and although everyone can hear, they cannot understand.
The key to successful PA (public address) in reverberant spaces, which includes places like train stations too, is not loudness, it is clarity. Clarity means that every word and every syllable can be heard and understood by everyone, even in the largest and most reverberant spaces.
The way to achieve clarity in a reverberant places of worship is not to use just two big speakers at the front, like an enormous hi-fi, or music concert. In contrast, many small speakers should be used, so that any individual member of the congregation is never more than around two or three meters from a speaker. None of the speakers needs to be very loud, and reverberation never has chance to build up.
In fact it is amazing at how good such a system can be. For places of worship where amplified music is a necessity, then a conventional music PA rig can be used to supplement the speech system, as long as the operators – usually volunteers – clearly understand what each system is for.