TA is NOT designed for triangulation because antenna is very long (about 300 meters on Perry and little shorter on submarines). So you just cannot measure TA contacts from one point.
TMA is only as good as you are good in this job. So if you arent supercomputer, just take a note: bigger ol lower errors exists ALWAYS on your solution. If you understand it sooner, you will be more effective in TMA.
Don't watch FPSchazly's TMA videos - he tries to do this too perfectly and only for AI moving straight targets. In MP this kind of "perfection" (all dots in perfect line) never happenes. Yuo should to learn how to use "unperfect" TMA solutions for predicting where is your target