"nobody knows to what extent the expansion of the system (DMS)...

"nobody knows to what extent the expansion of the system (DMS) will then influence / disrupt the operation of the system and even have to be shut down again for safety reasons."
Typically the biggest impacts of a change in systems are downstream. There is no downstream processing after the DMS because its the past part of the processing chain, however there may be cascade upstream impacts where a change in crushing, washing or screening will assist the DMS to work better.

At least in the smaller particle sizes used in lithium (sub 10mm), experts have noted DMS units work best when the material is a consistent size. This issue may not apply as the material gets larger. E25 is planning to use the DMS with an operational range of 8mm to 85mm. The upper end is increased from the existing 63mm so there is a favourable increase in upper crush size flexibility. There are likely to be some optimisation work as the E25 establish how clean ore needs to be. Working on density rather than colour, the DMS is likely to have greater tolerance for minor surface clay colours. This will be a favourable cascade effect up the chain as it may allow a faster throughput speed for the log washer. Hopefully E25 have the sense to experiment with creating size grade ore feeds so that for example over period X larger sizes are processed, over period Y smaller sizes are processed.

The DMS system proposed by E25 is the most simple option available because it only uses one DMS unit with the dense media set at 2.7x density (sinks = ore produce, floats = waste). In material where achieving an acceptable recovery rate is difficult DMS units operate in pair's (Primary/secondary). This allows the density ranges of float, middlings and sink. Middlings may be recrushed to a smaller size and go through another pair of DMS's. Some systems like that proposed by Atlantic lithium have 6 DMS units as 3 pair's for progressively smaller sizes. Its interesting that the flow chart doesn't have a smaller secondary DMS enabling middlings for recrushing. This might be something E25 look at once HPMSM is funded.

DMS units are not some newfangled invention that have unknown safety implications. They are a well understood piece of mining concentration equipment that have been used for decades. A quick google found references dating back to the 1930's and a 1942 patent. There will be additional site risks when site construction is underway, but that is no different to any other construction site around the country.

The picture below shows where the plant is proposed to be placed. It will alter traffic routes around the site but is primarily separate from other activities.