Computer Assisted Qualitative Data Analysis (CAQDAS) like NVivo, Atlas, etc. may have become the trend for qualitative studies. I still think that the reported limitations of CAQDAS (sometimes requiring 'data cleaning', etc.), would continually require significant manual process and efforts of qualitative data analysts, hence, the need for it to be undertaken by experienced persons.

Most CAQDAS programs build directly on the basic procedures of manual coding (although some are now experimenting with automated coding). For small datasets, there is little to be gained by using software versus hand coding. So, I do not believe that journals are ever likely to require the use of software.

Yes most CAQDAS programs build directly on the basic procedures of manual coding, some platforms like DiVoMiner® can do machine coding, but it is suggested to combine the two methods do improve the accuracy. And if you are doing quantitative content analysis, for example, doing inter-coder reliability test or doing descriptive statistics, DiVoMiner® is easy to use!