The answer IMO goes back to economics... supply & demand.    Once the Bitmark blockchain started flowing steadily with v0.9.7.0, after many years of producing  very few blocks per day,  many more coins are being produced than the market demand to absorb them at 10k sat. was used to.
 Also, particularly because a majority of blocks are being merge-mined, perhaps they are seen as "freebies" or extras by the merge miners, to be liquidated ASAP on exchanges.

While the project does not specifically target any price for the coin, one purpose of the Coin Emission Modulation (CEM) algo is to buffer, to some degree from these types of market pressures, by issuing coins according to demand, as sensed by network hashrate. Algorithmically, the implementation has worked well, but the limited action range of 50% epoch block reward has not proven to be strong enough. The current 50% CEM v0.1 action range will be expanded to at least 80%  in CEM v0.2.

Additionally, an intense debate has been had on the Bitcoin GitHub regarding the pros / cons of rewarding merge-mined blocks the same as native-mined blocks.    ( Merge Mining Debate )    The discussion ensued after it was raised as an Issue by GitHub user Jarr0s.

Merge-mined blocks are of near-zero marginal cost for merge miners to produce, because their fixed equipment costs do not change at all when they merge-mine MARKS. They are already in operation, mining another chain that shares one of Bitmark's 8 algos.  They just have to set up a Bitmark node on their existing hardware, and dedicate a some extra processor cycles to submit work for Bitmark.

I am of the opinion that it is appropriate to reward them at a fraction of native-mined blocks, because they are of much lower cost to produce than native blocks. Is it simply a principled question of Equity.

The project appreciates merge-miners. The blocks they mine, linked to higher hashrate parent chains, provide addtional security to the Bitmark blockchain, and have value for Bitmark and will be rewarded - just not at the same rate as native blocks.
If merge miners wish to get the higher native rate, they may mine MARKS natively, as the parent chain.

These measures are likely to result in  a coin emission rate more in line with the historical averages, and may restore the market conditions when the Bitmark price was in trading band an order of magnitude higher than it is now, but Disclaimer: this is just speculation on my part.      Market forces are always outside of the control of any one group or individual.

Our purpose and vision from the start has been to be a solid, secure and reliable blockchain on which to implement the myriad possibilites of Marking.      Bitmark v0.9.7 is a solid step in that direcition, because it definitely restored prompt transactions processing, with block production fairly steady at the target 2 minutes.  The idea of decoupling coin emission directly from block production proved feasible and succesful with CEM v0.1, and wil be built upon and strengthened in the next iteration.

Bitmark version 0.9.8 is being coded and will then enter testing. Once it is final, we will publish the exact specifications of CEM v0.2 and the policies regarding native and merged mined blocks.

I am eager to move past the blockchain dynamics and start focusing on the Marking, which is what will bring the best value to Bitmark.