If you want the stringers to maintain their shape and not integrate into the flattened glass, you need a tack fuse. The degree of tack fuse is the next question, i.e., how much of the shape to you want to maintain? (And if you tack fuse upside down, the base glass will flow down onto the shelf and fill in around your stringers, so tack fuses need to be done face up, thinfire or not)
Tack fuses can be tricky--all those shapes tend to try to contract away from each other as the glass cools, potentially creating stress points that might crack or pop off the glass. The more fused/flatter they are, the less likely that is to happen. The trick is to using a slower cooling schedule to prevent those stresses from forming so tack fuse schedules are almost always longer/slower than a flat fuse.
With a normal flat fuse, you measure the thickness of the glass at its thickest point and follow a schedule for that thickness. With tack fuses, you do the same measurement (except you measure to the top of the highest component instead of just the thicknesses of the layers), then you multiply it by some factor and use the schedule for THAT thickness. Generally, if the glass is mostly melted into the surface and just showing as rounded bumps (sort of like a sine wave), I use the schedule for a double thickness. If I want the tack fused glass to pretty much keep its shape and the edges to stay relatively sharp, I might opt for 4X the thickness.
Exactly what I do depends on a lot of things (and I tend to experiment first before working on the final piece). If there is a big contrast between the glasses (i.e., light-colored opaque glasses paired with black or dark transparents) or I'm using lots of colors and shapes, or there's a big variability in thickness from one spot to the next, I'll use a thicker/slower schedule. If the colors are similar and the thickness is pretty consistent, I can go faster. I do quite a bit of experimenting and I've written some stuff about tack fuses on my blog--if you check that might give you some ideas:
http://www.morganica.com/bloggery/2010/ ... -samplers/