Right now, the result objects from Rsubread::featureCounts can easily read into most DE Analysis packages (at least for DESeq2 and edgeR, which should cover most use cases). However, at this time, it appears to me that there is native support in parsing the output file(s) from the command-line version of featureCounts.

While I understand many of us would rather use the Rsubread version, I can still foresee usage cases where the command-line version is desirable (for example, when using a job scheduler--otherwise a wrapper Rscript is needed). So I think either one of these would be desirable:

1. Rsubread adds a function to read the output file into Rsubread::featureCounts objects. I suppose the code already exists in Rsubread::featureCounts? Or,
2. Adding an extra type to tximport::tximport to read the same files into a tximport object.

I have no preference over these two approaches, but maybe some other user may chime in on this.

You can read command-line output from featureCounts by

library(edgeR)
y <- readDGE( fcoutputfiles, columns=c(1,7) )
y$genes <- read.delim( fcoutputfiles[1], row.names=1 )[,-6]

BTW, if you are using Rsubread::featureCounts, there's a fairly recent function featureCounts2DGEList in the edgeR package that might be of interest. This function is particularly useful to handling more complex featureCounts output for exons and exon-junctions, but it is also useful just for simple bulk RNA-seq as well.