Text images are just arrays of numbers stored in a tab-delimited file (https://en.wikipedia.org/wiki/Tab-separated_values), where each location in the file is a pixel and the value stored there is the pixel intensity.
Many softwares (e.g. Microsoft excel) don’t support the saving of arrays as TIFF files. Many of these such softwares do support the saving of arrays as (tab-separated) text files. So, whether you like it or not, you might come across an image that was saved as a text file. You might not (if you’re lucky), but I have, so being able to read them is handy. Beware that ijtiff::read_txt_img() assumes a tab-separated file (so something else like a CSV file won’t work). This is the type of text image that you can save from ImageJ.
A 32-bit TIFF file can only hold values up to \(2^{32} - 1\), that’s approximately \(4 \times 10^9\). For whatever reason, this might not be enough for you, what if you want to write a value of \(10^{10}\) to an image? Then, you’re out of luck with TIFF files (and most if not all other image formats), but a text image is your friend. Text images place no restriction on the values therein. They’re awkward and inefficient, but they can get you out of a hole sometimes.