There was a definite issue with the match-up between engine and transmission. My understanding (and I'm no engineer) was that though individually they were sound equipment they did not "match up" very well and this caused problems.
Not "match up" as in "fit together", but matching in the power-transmission sense of impedance matching. The transmission matches the source impedance of the engine to the load impedance of the wheels, by selecting whichever of a given set of ratios is most appropriate for the conditions of speed and load. The "poor matching" problem is when the set of available ratios is badly chosen and speed/load combinations frequently arise for which none of them is very appropriate. In a car you might express this by saying, for instance, that the gearbox has a big gap between second and third, meaning that you have to rev the nuts off it in second and then chug it sluggishly in third as you accelerate, and it is more of a pain to drive than one where the ratios are less disparate.
With a torque converter transmission you don't have sets of fixed ratios, rather you have sets of bands of possible ratios, with the instantaneous ratio varying itself continuously and automatically within each band between the set limits; each band will have a particular point of greatest efficiency and become less efficient either side of that. The continuous variability gives you more accurate matching than is possible with a set of fixed ratios, and it is more tolerant of suboptimal selection of the ratio band limits, but because the efficiency curve is not flat the problem does still exist. The Westerns did have exactly the equivalent of a big gap between second and third, and the result was that their performance at speed was a bit disappointing compared to what had been expected and what the engine power would normally imply. It wasn't the sort of thing that made them go bang and stop pulling the train though.
There were rather a lot of awkward constraints on selecting the transmissions for the Westerns, for reasons such as the considerations of size, weight and alignment that caused difficulties with every large component in trying to get the overall design to fit the loading gauge and axle weight limits, and the shortage of suppliers considered suitable severely restricting the number of possibilities to choose from. It didn't become apparent until the design was well advanced that they had got into a bit of a hole and ended up with their options unduly restricted, so they were no longer able to select a transmission with as appropriate a set of ratios as they had intended.
Of course these days now that we can design torque converters using CFD instead of by hand...
As for "more depots to service them", you've got hold of the wrong end of the stick.
The point was Canton was set up as a single centre of excellence to work on 37s and 47s only and looked after the whole WR diesel electric fleet. No such centre of excellence was set up for the hydraulics, and repair work was scattered over OOC, Swindon, Bristol BR, Newton Abbot and Laira, often in dirty repurposed steam-era buildings. If a single hydraulic central depot had been set up, then maybe servicing would have been easier and costs reduced due to economy of scale
Of course there was an element of learning from experience going on there, but as I remember it the programme of setting up heavy mechanical facilities for servicing the hydraulics got booted in the gut through some funding/politics dropping, and ground slowly to a halt without ever providing the level of facilities that had been intended (Laira being late to get its engine lift facilities being one aspect of this). In particular the unit replacement system for engines and transmissions basically just never happened, so instead of whacking in a new lump and getting the loco back in service in a day or two while the one you'd taken out got repaired in a workshop built for the purpose, you'd have them sitting out of traffic for weeks while fitters guddled about swearing in the muck with everything in situ, and swearing even more when they couldn't get at things which according to the original idea they would never have had to get at in situ anyway. (Not to mention the supposed interchangeability of engines and transmissions coming apart because someone forgot where the bolt holes were supposed to go, but that was more by way of icing on the cake.) You could argue that since they nevertheless
did manage to achieve availability figures of the same order as diesel-electrics and considerably better than some which outlasted them in the end, it's a point in their favour...