It seems like a very difficult problem to address, you are probably better off comparing the counts at the TSS compared to some range downstream... (e.g. +1 to +5000) I would take a look at the data to get an idea for where to set this range. Maybe you could calculate this ratio for every gene in each condition and then compare it between conditions.
The rationale for decay-rates is that the intron-exon coverage should be relatively constant (imagine some kind of steady state) but when that system is perturbed the rate of transcription in the nucleus will decrease and by proxy the amount of intronic reads. The same goes for mRNA decay. This oddly intrinsicly linked to the idea of RNA-velocity.
[introns] + [exons] = total concentration
if [introns] goes up and [exons] goes up or is constant it suggests transcription is upregulated
if [introns] stay constant and [exons] go down it suggests increased decay rate
if [introns] go down and [exons] stay constant it suggests transcription is downregulated
if [introns] stay constant and [exons] increase it suggests decreased decay rate
All of these pathways are intertwined so the assumption is actually not ideal... and probably not as thoroughly investigated as it should be.