I just tried the LastMAF import mode with rRNA data from NCBI. (Thank you for implementing it!)
Unfortunately, it does not seem to take into account that LAST does not order output, so reads get multiple assignments. I tried running LAST in a single thread (or at least parallel-fasta using a single process), but this didn’t help.
As an example, LastMAF discovered 112,315 reads in the input, whereas BlastTab, using the same input MAF but ordered and reformated to look like Blast output (see my reply to another thread here), discovered only 9,612 reads.