Jitter and optimising fmri task design

Hi Afni Team

My first question is a bit of basic one about jittering. I hope you guys can help me. I have tried to find more info on this type of design but to no avail, so any guidance on this would be great!

Is it a problem to jitter the length of events within a trial rather than the ISI between those events? Here is an example of this type of design (see Figure 1):
plos paper

So instead of having jittered ISIs with a fixation cross or something, the events would not be fixed but would themselves be jittered in length. The reason I am interested in this is that it allows me to separate the events that I am interested in without having to include lots of ISI’s between events. This way I can include more trials. Plus subjects won’t have their working memories strained too much.

Another question, which is more afni-centric. If I did run this type of design, would it be straightforward for 3dDeconvolve to handle the varied lengths of the events?

Thanks a lot in advance for any help here!