So you mean they don't factor those into their plans? 
I am being deliberately provocative and I appreciate that things like that might happen in individual builds (although unplanned digging up roads will be a rare occurrence). I think my real point is that better planning, better management,
The problem here is an understanding of scale.
By "better management", you really mean that BT ought to be able to
project manage better, to get one cabinet (*your* cabinet) up as fast as possible, and to not slip. You then kinda "hope" that they apply the same measure to every other cabinet around... and Bob's your uncle... project done.
But BT's yardstick is different. They don't want any one cabinet to be finished as fast as possible. They don't even want *all* cabinets done as quickly as possible. They want 66% of the population done by (some point in) 2014 but most importantly they want it done to budget.
The usual ways to get a programme done *quickly* are to stack the sequence of activities as close together as possible, and to have as many people work on it as possible, in parallel.
The problem is that any delay anywhere will result in people sitting idle, waiting to be assigned new work. And there goes the budget...
If the project is to run *cheaply* then you don't try to run each activity too close together - so you don't even request a power connection until you know the cab is in place, for example. It means the installation time for any one cabinet is extended by periods of sitting idle, and the installation time for a cabinet with problems has those idle times amplified. But the *people* are kept fully busy.
When you own 80,000 cabinets, and need to enable 200 every week, you are soon going to find a percentage of delays. As BT know the *average* install time, and the average number that encounter delays, they know how many cabinets have to be being installed in parallel - taking into account both wasted leadtime *and* wasted delay time. They can start the work for a lot more cabinets, knowing that delays *will* be encountered, but without knowing *where*.
The overall result is that the national plan is well ahead of schedule (the target date has moved forwards from late 2015 to mid 2014), but the schedule of a few individual cabinets can vary hugely.
The fact is that BT appear to be managing the project rather well - from their perspective. But individuals may see something different - and those poor souls who connect via the very few of the *really tricky* cabinets get shafted because...
and most importantly, better communication of what's going on and what's going wrong
This aspect is what they've been really bad at.