If trying to provide information and possible reasons is defensive? Given the tone of many of the posts is it any surprise? Industrial action? Really?!?
The tower doesn't care less about Maestro - it's just a tool for providing an ordered stream of arriving aircraft and they're not fussed about how it's achieved. The idea is Maestro puts aircraft in more or less the right place but arrivals tweaks the sequence to an extent and TMA tweaks it further - some for separation and some to account for the dynamics of the aircraft around you. Sometimes Maestro's wind and performance model isn't so crash hot and you get dead heats from different direction. I'll emphasise this - Maestro is a sequencing tool at the threshold and not a separation tool.
How it works: Maestro calculates your estimate for the feeder fix based on winds, levels in your flight plan (and as altered in-flight either automatically or manually as the auto doesn't work) and the TAS in your plan (and as updated manually by us on advice from you). It then calculates a time from the fix to the threshold via the STAR based on winds and performance data (based on historical data collected for type and airline). This sets an order but as your fix time is periodically recalculated by Maestro the order isn't fixed yet - this stage is called "unstable" as your position in the sequence isn't set.
The flow sets the acceptance rate - however many seconds between arrivals based on wind (your groundspeed is lower down final with stronger wind) the mode of ops (biased towards arrivals or departures) and type of approaches being used. No. 1 goes through untouched. No. 2 gets a fix time calculated (if necessary) based on the acceptance rate and 250kts from the fix. No. 3 gets a fix time to fit it the required time behind no. 2, and so forth. If there are several aircraft with untouched landing times within a few seconds the the sequence can jump around a lot - you might go from no. 1 to no. 5, so from no delay to 10 minutes (or whatever).
The next stage is "stable" where the sequence can still change but usually not with out manual intervention. Left to its own devices Maestro will make you "stable" 15 minutes from the fix - it won't update your untouched fix estimate further. Generally we'll set this quite a bit earlier to stabilise the sequence - we'll set an untouched fix estimate based on what we think you'll achieve (based on observed performance) or what we think you can achieve with track shortening and high speed if there's a gap. Once stable Maestro won't automatically change your untouched fix estimate but the order can still be changed by a close in departure getting away and being added to the sequence, or if we manually change it. Also blocks can be added to accommodate things like medical traffic into Essendon if needed or adding extra spacing behind A380s. Your feeder fix time can also change if an aircraft has their untouched fix time adjusted, placing them ahead of your untouched time.
There are two other phases, super-stable and frozen but they don't really have any further impact on times changing, other than for instance putting a go-around back in the sequence, but that's all manually done by the flow. FWIW we also have to insert Essendon traffic into the Melbourne sequence when vis is poor, although that's not too often. Adding a slot for Essendon medevac traffic is far more common.
Whether you are issued a time and vectors to lose the rest (if necessary) or a hold depends on how long the delays are going on for - at some stage the delays just get too large to absorb sensibly with speed reduction and vectors from a controller workload perspective. Maestro has reduced the amount of holding we do and the ground delay program has reduced it further. I can't quantify this but I certainly do less holding now than I have in the past.