For recognising limitations of QPN-Tool a quick glance at similar and well-known existing tools might be helpful: GreatSPN  from the Universita di Torino, Italy, and SPNP  from Duke University, Durham, USA.
GreatSPN supports the specification and analysis of GSPNs and DSPNs. It provides a graphical interface which allows the modelling of only uncoloured nets but including inhibitor arcs, marking dependent rates and probabilities. Difficulties in describing non-trivial queueing situations and scheduling strategies as mentioned in Sect. 1 occur. GreatSPN offers an ample variety of algorithms for qualitative analysis which contains reachability graph analysis, computation of invariants, deadlocks, and traps, and the (inverse) token game. Net properties are nicely animated on its graphical representation. For quantitative analysis, Markov-chain based transient and steady-state analysis is provided as well as simulation. Output measures have to be textually defined. Compared to GreatSPN, future versions of QPN-Tool should be able to handle inhibitor arcs, marking dependent rates and probabilities, and deterministic times. Analysis algorithms of QPN-Tool do not contain the token game, transient analysis and simulation yet.
SPNP is based on the analysis of Markov reward models. Its textual interface is closely related to the programming language 'C', the language SPNP is implemented in. It allows marking dependent arcs, marking dependent enabling functions and general priorities. According to its descriptive power only reachability-graph based qualitative analysis is performed. Its main focus is on quantitative analysis based on Markov reward models (transient and steady state analysis). Output measures are specified by user-defined C-functions supported by a set of predefined functions. An automated sensitivity analysis is offered which derives different CTMCs from a fixed state space by variation of an independent parameter for firing rates and probabilities. Compared to QPN-Tool the modelling process in SPNP tends to be a programming process with a strict focus on Markov reward process analysis and few qualitative analysis features as a debugging aid. A variety of qualitative analysis algorithms like in QPN-Tool or GreatSPN is not given in SPNP. As far as quantitative analysis is concerned, SPNP differs from QPN-Tool by its transient analysis, automated sensitivity analysis and its ability to handle general reward specifications.
This comparison is not supposed to be exhaustive or to give a complete characterisation of GreatSPN and SPNP. We just wanted to demonstrate limits of QPN-Tool to draw the following conclusions.