Backfiring in the exhaust can happen in any gear and depends only on engine load which increases with engine breaking. So you can get it with any downshift or rapid deceleration. This results from a lean condition where the gas/air mixture ignites in the exhaust rather than in the cylinder. I had this problem big time with a full Jardine exhaust and triple filters. I had a PCIII installed and had it dyno tuned. The tune was great except it exploded, not just popped. I noticed the tuner put large negative trims (less fuel) at zero throttle. I've heard this strategy for getting rid of popping but it didn't work for me. Most tuners and Dynojet, recommend adding fuel at zero TP. So I added 15% more fuel from 1700 to 4700 RPM at zero TP and the popping went away.
The L tables supply fuel as a function of manifold pressure (MAP) and rpm for low throttle position (TP) which for the stock tor tune (20355) is below 31% between 2521 and 3775 rpm. Backfiring occurs at zero throttle, so it's using the L tables. To get rid of the popping, you have to add some fuel in the L table columns corresponding to the MAP pressure and RPMs where the popping occurs, which you don't know unless you can log it, at zero throttle. Comparing 20355 and 20352 (stock exhaust), you can see more fuel in some cells of the L tables in the low pressure (220 mb) columns in the 20355 tune, so Triumph tried to address this. The F tables are for higher TP and supply fuel based on TP versus RPM. These are called the main fuel tables.
You should also make sure there are no leaks at any of the exhaust connections. This can cause decel popping.
So, it looks like Hanso's tune got rid of the popping but there are other issues. I'm sure Hanso's tune derestricts the power by opening the secondaries 100% at all TP and RPM. Also, he probably disables the O2 sensor, which has the good effect of reducing the decel popping but could have other effects like less smoothness at low TP. I've switched back and forth a million times. During winter, when it ran rich, I'd enable the O2 sensor for smoothness and because it was richer anyway since it's freezing here, and it didn't pop as much. In summer, it ran leaner because of the extreme heat and I'd disable the O2 sensor to reduce popping.
In the stock tunes, the O2 sensor is always enabled so it's likely to adapt and lean the fuel out at low engine load. This causes popping. You're probably aware that the later model R3Rs are "derestricted". All they did was open the secondaries 100% at all loads except at the high end. You can see this by comparing your tune (20355) with the later 20776 for tors. The only difference between these tunes is in the secondaries table. The fueling and timing are identical. All you need to do to derestrict the stock 20355, according to Triumph's definition of derestrict is to copy the secondaries table from 20776 into 20355. You might try this just to see how it runs.
And: check for leaks.